Cost Optimization of Hybrid Islanded Microgrid for

Rural Electrification
Rabia Khan Noel N. Schulz
School of Electrical Engineering and School of Electrical Engineering and
Computer Science Computer Science
Washington State University Washington State University
Pullman, WA 99163 Pullman, WA 99163
Email: [Link]@[Link] Email: [Link]@[Link]

Abstract—Energy poverty is a big enigma, that needs to be to consider low carbon technologies for future deployments of
considered for millions of people suffering in remote rural power generation. During the past few decades, penetration of
communities without electricity access. As the grid extension distributed energy resources (DER) in the grid has increased
is cost prohibitive, the renewable energy sources (RES) are
an optimal solution for off-grid communities. In this paper, a to decrease the usage of fossil fuels [5].
hybrid microgrid system is designed and simulated in Hybrid The major challenge today is to provide electricity to
Optimization of Multiple Energy Resources (HOMER) based on the un-electrified people where the grid has not reached
the estimated load demand and available energy resources. Three yet. There are three possible solutions. 1) Grid extension,
different rural villages from Ethiopia, Uganda, and Brazil are 2) Standalone Solar Home Systems (SHS) and 3) Islanded
considered for a comparative performance analysis based on the
least cost of energy (LCOE). Two different dispatch strategies, Microgrid Systems. Grid extension is a reliable but much
i.e., i) Load following and ii) Cycle charging are modeled with expensive solution for providing electricity access to the
the different scheduled operation of the diesel generator and far-off diverse clusters of the population. Secondly, the losses
bio generator. The solar and wind resource data are obtained associated with long-distance power lines make this strategy
from the National Solar Radiation Database (prediction of the ineffective. The second option, SHS, have been incorporated
worldwide database) [1] and the National Renewable Energy
Laboratories (NREL) [2] respectively. The results of sensitivity in rural areas. It is a low upfront cost solution but has limited
analysis, techno-economic analysis and cost optimization based electrification capabilities, e.g., scalability, flexibility, and
on LCOE verify the feasibility of the proposed microgrid system. limitation to support large loads [6]. In comparison to the
first two technologies, islanded microgrids are considered
Index Terms—Rural Electrification, Renewable Energy Sources, to be a better solution due to their capability of supporting
HOMER, Off-grid, Cost-Optimization
electrification beyond substance level living. The RES based
islanded hybrid microgrids provide a competitive solution
I. I NTRODUCTION
to the other solutions particularly for the energy stricken
Lack of access to electricity is an important electricity crisis communities [7].
for people living in rural areas. According to International
Energy Agency (IEA), the number of people without any The off-grid communities highly depend on conven-
access to electricity worldwide has dropped down to less than tional energy resources like kerosene, natural gas, candles,
one billion for the first time in 2017, but people without access petroleum, coal, and animal dung to meet their energy needs.
to clean cooking facilities is near 2.7 billion globally [3]. Most These resources have harmful effects on human health and the
of the population lives in rural countries of Sub-Saharan Africa environment. Hybrid microgrids serve as a viable solution to
and South Asia. Energy poverty is a severe issue that needs to provide electricity to the communities in developing countries
be addressed for the sustainable development of a community. [8]. The current international trend in rural electrification is
Electricity access is highly related to the human development to utilize the available renewable energy resources based on
factors, i.e., healthcare facilities, sanitation system, educational the location. Solar photovoltaics (SPV) and biomass energy
access, and rate of employment [4]. systems are considered ideal because of the abundant solar
The increase in the world’s population during the past radiations and the significant biomass availability [9].
decades has resulted in excess usage of fossil fuels. This Several authors have designed and studied the performance
increased consumption of fossil fuels has led to their high of microgrids with different sources and loads. Gelma et.
rate of depletion [4]. Fossil fuels have a harmful effect on the al [10], Samuel et. al [11] and Geetachew et. al [12] have
atmosphere. They are considered to be a major cause of global designed solar/PV/wind, PV/wind/diesel, and small-scale Hy-
warming and drastic changes in climate. The power generation dro/PV/Wind based hybrid microgrid system for Ethiopia
technologies employing fossil fuels cause harmful emissions and considered their feasibilities. However, their study lacks
that have an adverse effect on the environment. It is imperative the cost-optimization and sensitivity analysis of the designed

978-1-7281-1981-6/19/$31.00
Authorized licensed use limited to: Thapar ©2019
Institute ofIEEE
Engineering & Technology. Downloaded on April 22,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 TABLE I: Cost of the designed microgrid components
Replacement Life time
Components Initial Cost O&M Cost
Cost
($/kWh) ($/kWh) ($/kWh)
($/kWh) (years)
SPV 3500 3148 35 25
Wind Turbine 30000 25000 50 20
Diesel Gen 2500 2000 14 25
Bio Gen 2200 1700 10 20
Converters 350 350 25 10
Batteries 900 800 15 8

microgrid systems. In this paper, the sensitivity analysis,

techno-economic analysis, and optimization are performed
using HOMER based on LCOE.
In this paper, a hybrid microgrid system is designed and Fig. 1: Schematic Design of Hybrid Microgrid System
simulated in Hybrid Optimization of Multiple Energy Re-
sources (HOMER). The cost optimization of the designed
microgrid system is performed based on the least cost of design. They include Dembile, Sao Tome and kampala from
energy (LCOE). Three different rural villages are considered Ethiopia, Brazil, and Uganda. The geographical coordinates
for a comparative performance analysis. People living in rural and time zones of the mentioned areas are given in Table. II.
areas have very low-income so the cost of electricity is the
most important factor while designing the microgrid system. TABLE II: Geographical coordinates of the locations
Section II presents the design and modeling of hybrid micro- Location Longitude Latitude Time Zone
grid system in HOMER. In section III, performance analysis Dembile, Ethiopia 6◦ 0’ North 37◦ 56’ East GMT + 3
of the designed microgrid system is illustrated with different Sao Tome, Brazil 52◦ 59’ West 23◦ 59’ South GMT + 3
Kampala, Uganda 0◦ 34’ North 32◦ 58’ East GMT + 3
dispatch strategies and scheduled operation of resources em-
ployed. Section IV presents the optimization results for the
case studies. Finally, in section V the conclusion and future
scope of the designed system and methodology is presented.
II. S YSTEM D ESIGN AND M ETHODOLOGY
The three different locations from the world’s poorest
countries with respect to electricity are chosen to evaluate the
feasibility of the designed hybrid microgrid system. Based on
the estimated load demand and resources available, an optimal
hybrid model with the least cost of energy is selected for each
location.
A. Proposed Hybrid Microgrid System
HOMER is a Hybrid Optimization Model for Electric
Renewables used to design and simulate the microgrid system
based on the components selected and the locations selected. Fig. 2: Load Profile for DC, AC, and deferrable loads
Homer performs energy balance calculation and gives the
optimized solution based on the parameters opted. In this
C. Load Estimation
paper, a hybrid microgrid system consisting of solar photo-
voltaics (SPV), wind turbines, diesel generators, batteries, and Three kinds of loads, i.e., DC, AC, and deferrable loads are
bio-generators are designed for three rural areas of Ethiopia, assumed based on the type of appliances, energy usage based
Brazil, and Uganda as shown in Fig. 1. The load is subdivided on activities and daily routine of inhabitants. DC loads consist
into three types and their profiles are estimated. The proposed of electric lights, e.g., lamps, tape recorder, Razor, charging
hybrid microgrid system is evaluated by sensitivity analysis, batteries etc. AC loads comprise of refrigerators, health center
cost-optimization, and techno-economical analysis. HOMER and school loads. The DC and AC loads are assumed to be pri-
displays a list of configurations for the hybrid microgrid mary loads while the water pumps, harvesting machines, and
system based on the LCOE. The cost of components used grinding machines are secondary loads and are classified as
for designing the microgrid are listed in Table. I. deferrable loads. The load profile for DC, AC, and deferrable
loads is estimated from the past projects [13]–[15] while the
B. Study Area hourly load demand is obtained by creating an excel sheet
The villages with no electricity access are selected for ana- based on the knowledge of the load demand and the duration of
lyzing the most optimal hybrid stand-alone microgrid system consumption. For example, lighting loads are mostly switched

Authorized licensed use limited to: Thapar Institute of Engineering & Technology. Downloaded on April 22,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
on at night while the harvesting mills, schools during the day,
whereas the entertainment devices on weekends. The average
consumption of DC, AC, and deferrable loads is 260 kWh/day,
52 kWh/day and 7.5kWh/day while their respective peak loads
are 57 kW, 8.1 kW, and 750 W respectively. The load profiles
for the designed microgrid configuration are given in Fig. 2.

D. Resource Assessment
The availability of energy resources is entirely dependent
on the geographical coordinates of the location. Most of
the hybrid microgrids designed for the rural communities of
Ethiopia, Brazil, and Uganda were based on solar photovoltaic
(SPV), wind and biomass. In our designed model, the purpose
is to compare the different topologies based on their cost and Fig. 4: Monthly wind speed data for Uganda, Ethiopia, and
get the most cost-optimized hybrid microgrid model for the Brazil
three locations. Additionally, the sensitivity analysis is per-
formed to obtain optimal microgrid system configurations for
the sensitivity variables chosen. In this paper, three sensitivity 3) Biomass Energy resource: The main source of biomass
variables are chosen. They include DC load, AC load, and as a fuel in rural areas are the crop husk and the livestock
Global Solar radiations. waste. Due to extreme poverty in these areas, the crop’s residue
is mostly fed to the cattle and donkeys so the livestock waste
1) Solar Energy resource: The solar energy resource po-
(animal dung) is used as a biomass energy resource for the
tential for the selected locations is calculated by using the
villages. According to the survey in [9], the amount of average
geographical coordinates, i.e., longitude and latitude. NASA
animal dung is calculated for each month by considering the
Surface Metrology and the National Solar Radiation Database
availability of feed intake and estimated lives of cattle. The
(prediction of the worldwide database) [1] is used to calculate
estimated values for the available biomass resources is given
the daily radiation (kWh/m2 /day) and the clearness index. The
in Fig. 5.
average solar radiation for SPV is 5.5 kWh/m2 /day, which is
the case for our chosen locations. The solar radiation data for
Ethiopia, Brazil, and Uganda is given in Fig. 3. III. O PTIMIZATION R ESULTS

The microgrid system design shown in Fig. 1 is optimized

based on the LCOE. The various sensitivity variables are
chosen for which the optimization process is repeated. The
sensitivity variables selected are DC load, AC load, and Global
Solar radiations. The different case studies are performed
based on the scheduled operation of the bio generator and
diesel generator. The bio generator’s operation is varied based
on the biomass feed (animal dung) available so the bio
generator is scheduled to operate for the time period as show
in Fig. 6. The diesel generator operates in case of emergency
or when the resource availability is limited. In case study 1
all resources operate in optimal mode, In case studies 2 and
3, the bio generator and the diesel generator are scheduled to
operate as shown in Fig. 6.
Fig. 3: Daily Radiation and Clearness index for Uganda,
Ethiopia, and Brazil

2) Wind Energy resource: The wind energy resource profile

is estimated from the National Renewable Energy Laboratories
(NREL) [2]. The procedure is the same as for solar energy
resource assessment, i.e., using the geographical coordinate of
our desired location and estimate the average wind speed (m/s)
for the whole year. The ideal wind speed for a wind turbine
to operate effectively is 4 m/s. The wind data profiles for our Fig. 5: Biomass Resource Data for Uganda, Ethiopia, and
selected locations are given in Fig. 4. Brazil

Authorized licensed use limited to: Thapar Institute of Engineering & Technology. Downloaded on April 22,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 TABLE IV: Cost-Optimization results for Ethiopia
Cost of
Case Dispatch
Optimized Model Energy
Study Strategy
($/kWh)
1 SPV, BioGen, Conv, Storage LF 0.019
1 SPV, BioGen, Storage CC 0.030
2 SPV, BioGen, Conv, Storage CC 0.025
2 SPV, BioGen,Conv,Diesel Gen,Storage LF 0.035
3 SPV, BioGen, Conv LF 0.076
3 BioGen, Conv, Diesel Gen CC 0.077

TABLE V: Cost-Optimization results for Brazil

Cost of
Case Dispatch
Optimized Model Energy
Study Strategy
($/kWh)
1 SPV,BioGen,Conv,Storage LF 0.019
1 BioGen,Diesel Gen,Storage CC 0.026
2 SPV, BioGen, Conv,Storage LF 0.020
2 SPV, BioGen,Conv,Storage CC 0.031
3 SPV,Conv, Storage CC 0.084
3 SPV,BioGen,Conv,Diesel Gen, Storage LF 0.086

Fig. 6: Operation of Resources

TABLE III: Cost-Optimization results for Uganda include DC load, AC load, and Deferrable load. Additionally,
considering the energy resources, the number of batteries,
Cost of SPV ratings, the price of diesel ($/L), wind turbines, and the
Case Dispatch
Optimized Model Energy
Study Strategy biomass resources can also be used as sensitivity variables.
($/kWh)
1 SPV, BioGen, Conv, Storage LF 0.019 The optimization results for the case studies 1,2, and 3 of
1 SPV, BioGen, Storage CC 0.026 Ethiopia are given in Fig. 7, Fig. 8, and Fig. 9 respectively.
2 SPV, BioGen, Conv LF 0.020
2 SPV,BioGen,Conv CC 0.031 The optimized model for each case study and it’s respective
3 SPV, Conv, Storage CC 0.084 dispatch strategy is given in Table. IV. The optimization results
3 SPV, BioGen, Conv, Storage LF 0.086 for Uganda are given in Table. III and shown in Fig. 10,
Fig. 11, and Fig. 12 for case studies 1,2, and 3 respectively.
The optimization results for Brazil are displayed in Table. V.
The operation of the microgrid system designed in HOMER The results show that the cost optimized model with respect
is based on two main principles, i.e., 1) Dispatch strategy to LCOE consists of SPV, BioGen, Converters (Conv) and
and 2) Energy management unit. The dispatch strategy is the Storage for all the locations.
allocation of resources based on the load condition. In the
case of hybrid microgrid systems, a control algorithm defines IV. C ONCLUSION
the optimal interaction of resources to meet the load demand Rural electrification in the remote regions of sub-Saharan
with the least cost of energy [16]. In this paper, each case Africa and Asia through grid extension is not economically
study for the designed system is simulated for the two dispatch feasible because of the low load demand, dispersed popula-
strategies, i.e., load following (LF) and Cycle charging (CC). tion, and difficult geographic terrain. In this paper, a hybrid
The dispatch strategy is a control algorithm which determines microgrid system is designed and simulated using the Hybrid
the flow of energy from energy resources to the loads and the Optimization of Multiple Energy Resources (HOMER). Three
storage based on the priority chosen and scheduled operation villages from the world’s most energy-deprived countries, i.e.,
of resources. The optimization of a system’s performance can Ethiopia, Uganda, and Brazil are considered for evaluating the
be done by using the appropriate dispatch strategy, which performance of the designed microgrid. Comparative analysis
depends on a number of factors. They include ratings of among them is done by performing the sensitivity analysis,
the generators and storage, fuel price, load demand and the techno-economical analysis and cost-optimization using two
hours of operation for the load. In load following strategy, dispatch strategies (Load following and Cycle charging) for
a generator is run to meet only the desired load demand. the designed microgrid. The operation of the energy sources
Load following is an optimal dispatch strategy for those hybrid is scheduled based on the most optimal timings and the
systems that include a number of RES. However, the cyclic available renewable sources. LCOE is used for deciding the
charging dispatch strategy requires the generator to operate at most optimal system type for the case studies.
its maximum capacity, even to charge the batteries.
R EFERENCES
A. Sensitivity Analysis
[1] [Link]
In rural areas, there is a low load demand which varies on a [2] [Link]
time basis. The sensitivity variables chosen in our case studies [3] Outlook EA. From Poverty to Prosperity. IEA: Paris, France. 2017:144

Authorized licensed use limited to: Thapar Institute of Engineering & Technology. Downloaded on April 22,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Fig. 10: Sensitivity results of Case Study 1 in Uganda
Fig. 7: Sensitivity results of Case Study 1 in Ethiopia

Fig. 11: Sensitivity results of Case Study 2 in Uganda

Fig. 8: Sensitivity results of Case Study 2 in Ethiopia

Fig. 12: Sensitivity results of Case Study 3 in Uganda

Fig. 9: Sensitivity results of Case Study 3 in Ethiopia

generation system for Ethiopian remote area. energy Procedia. 2012 Jan
[4] Patel AM, Singal SK. Off-grid rural electrification using integrated renew- 1;14:1760-5.
able energy system. InPower India International Conference (PIICON), [11] J. Aanchal and G. Samreet, ”Feasibility study of hybrid power plant,”
2016 IEEE 7th 2016 Nov 25 (pp. 1-5). IEEE. International Journal of Recent Trends in Engineering and Research, vol.
[5] Khan R, Mehrizi-Sani A. Comparison of fault current limitation with 2, no. 9, pp 172-178, 2016
saturable reactor and dynamic voltage restorer. InPower Energy Society [12] Tesema S, Bekele G. Resource assessment and optimization study of
General Meeting, 2017 IEEE 2017 Jul 16 (pp. 1-5). IEEE. efficient type hybrid power system for electrification of rural district in
[6] Nasir M, Zaffar NA, Khan HA. Analysis on central and distributed archi- Ethiopia. Int J Energy Power Eng. 2014;3(6):331-8.
tectures of solar powered DC microgrids. InPower Systems Conference [13] Sepulveda TT, Martinez L. Optimization of a Hybrid Energy System
(PSC), 2016 Clemson University 2016 Mar 8 (pp. 1-6). IEEE. for an Isolated Community in Brazil. Int. J. Renew. Energy Res. IJRER.
[7] [Link] 2016 Dec 31;6(4):1476-81
com/CF4FnpPQ15q2 [14] Tucho GT, Nonhebel S. Alternative energy supply system to a rural
[8] Hirsch A, Parag Y, Guerrero J. Microgrids: A review of technologies, village in Ethiopia. Energy, Sustainability and Society. 2017 Dec;7(1):33
key drivers, and outstanding issues. Renewable and Sustainable Energy [15] Komakech-Akena A. A techno-economic Analysis of Grid extension
Reviews. 2018 Jul 31;90:402-11. against Stand-Alone Generation Systems in Sezibwa, Uganda. Mini-grid
[9] Hailu EA, Mezgebu C. Design and Simulation of Standalone Hybrid and Hybrid Systems.:1
(Solar/Biomass) Electricity Generation System for a Rural Village in [16] Girma Z. Hybrid renewable energy design for rural electrification in
Ethiopia. Ethiopia. Journal of Energy Technologies and Policy. 2013;3(13):38-52.
[10] Bekele G, Boneya G. Design of a photovoltaic-wind hybrid power

Authorized licensed use limited to: Thapar Institute of Engineering & Technology. Downloaded on April 22,2024 at [Link] UTC from IEEE Xplore. Restrictions apply.