Energy 161 (2018) 1211e1225

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Study on the operation optimization of an isolated island microgrid

with renewable energy layout planning
Shin'ya Obara a, *, Katsuaki Sato b, Yuta Utsugi b
a
Power Engineering Lab., Dep. of Electrical and Electronic Engineering, Kitami Institute of Technology, Koen-cho 165, Kitami, Hokkaido 090-8507, Japan
b
Power Engineering Lab., Dep. Electrical and Electronic Engineering, Graduate School of Kitami Institute of Technology, Japan

a r t i c l e i n f o a b s t r a c t

Article history: In this study, the planned facilities for the Teuri and Yagishiri Islands interconnection microgrid (small-
Received 10 April 2017 scale power network) systems were surveyed. In the small and isolated islands of this area, the wind
Received in revised form state changes dramatically with location and influences the production of electricity by wind power
30 June 2018
generators and solar cells. Based on the estimated values of the electrical power output by renewable
Accepted 17 July 2018
Available online 20 July 2018
energy, a method of optimizing the operation of the main electrical and heating equipment was
developed using a genetic algorithm (GA). The optimum conditions for installing the wind power gen-
erators include wind blowing from all directions at high speed throughout the year. Operation of the
Keywords:
Renewable energy
photovoltaic power plant is favored by high winds, which produce a large cooling effect over the solar
Electrical power cells. Adoption of the proposed analysis method will optimize the operation plan of the isolated island
Layout planning microgrid with a large utilization rate of renewable energy. The proposed operation plan obtained the
Operation planning effect of a planned utilization rate of annual renewable energy of 40% or more. The large rate of
Microgrid renewable energy is taken as the proposed microgrid without a battery, and the high effect was obtained
by the proposal design method of this study.
© 2018 Published by Elsevier Ltd.

1. Introduction utilization of local-supplyelocal-consumption energy on isolated

islands continues to develop. Hemmatpour et al. were proposed as
On an isolated island, which is not connected with a large-scale an operational tool for improving load ability of islanded micro-
commercial electrical power network, there are many examples of grids using an Adaptive Multi Objective Harmony Search Algorithm
independent power supply by diesel engine generators. Although a [1]. Neves et al. applied a microgrid accompanied by weather
diesel engine generator can supply stable electrical power, exhaust prediction of solar radiation to the isolated Corvo Island [2]. Meh-
gas elimination, and fossil fuel transportation are significant issues. rasa developed a control technique that enhances microgrids sta-
On the other hand, a clean energy system has been recently bility during the grid-connectedand islanded modes [3].
developed and applied that requires a distributed power supply. As Moreover, in Teuri and Yagishiri Islands (Haboro-cho, Hok-
a result, introducing renewable energy in such isolated areas is now kaido), the eco-island concept demonstration project, which was an
a major consideration. Teuri and Yagishiri Islands offer an example environmentally friendly endeavor, aimed at strengthening the
of local-supplyelocal-consumption energy use. In this area, previ- community infrastructure against disasters [7]. In order to support
ous studies were conducted on the load capacity and minimization this project, the study examined a microgrid (small-scale power
of electric power loss [1]; a prediction system for wind power and network) suitable for Teuri and Yagishiri Islands. In an island
photovoltaic production [2]; and the control required for stable microgrid, study on an economical energy supply which consists of
operation [3]. In Japan, such a microgrid system is applied on iso- a diesel generator, storage devices, and renewable sources to rural
lated islands in Kyushu and Okinawa [4]. That system consists of a areas [8], introductions of a grid-connected island microgrid in
battery and renewable energy source [5] as well as technology to China, Luxi Microgrid, with a flexible system structure and a hier-
control the fluctuations in wind power generation [6]. In fact, the archical control framework [9], hydrogen production by a renew-
able energy installed in an isolated island, and energy supply using
fuel cell power generation [10], design of single phase inverter
* Corresponding author. operating in island mode in a microgrid [11] and management of an
E-mail address: obara@mail.kitami-it.ac.jp (S. Obara).

https://doi.org/10.1016/j.energy.2018.07.109
0360-5442/© 2018 Published by Elsevier Ltd.
1212 S. Obara et al. / Energy 161 (2018) 1211e1225

islanded medium-voltage (MV) microgrid placed in Dongao Island

[12] are investigated.
On the other hand, there are considerably few examples about
the small island microgrid of cold regions. In the isolated islands of
a cold, snowy area, the period of high heat-to-power ratio is long,
and a significant amount of fossil fuel is required for heating.
Therefore, a heat pump system with a heat storage tank, running on
electrical power produced from renewable energy (photovoltaics
and wind power generators), is introduced, and an isolated island
microgrid with a high renewable energy utilization rate is planned
with fossil fuel consumption by an engine generator for electrical
power compensation (backup power supply). The wind state
changes significantly between different areas on these small and
isolated islands. Although the wind state affects the electricity
production rate of a wind power generator, the cooling effect of
wind affects a solar cell's skin temperature. Accordingly, when
renewable energy is introduced using a photovoltaic power station
and wind power generators, it is necessary to determine the loca-
tions on the islands with suitable wind conditions. Therefore, in
this study, we determined the installation locations for a photo-
voltaic power station and wind power generators through a wind
state analysis using a 3D topographic map. Based on the anticipated
electric power output by the renewable energy sources obtained
from this analysis, the installed capacity and cost of the electric and Fig. 1. Teuri Island and Yagishiri Island.
heating equipment were determined for the Teuri and Yagishiri
Islands interconnection microgrid system. Based on the estimated
When the renewable energy equipment is installed, depending on
values of the electrical power output by the renewable energy
the wind state, an equal amount of renewable energy will be sup-
installed at these locations, the operation method of “the main
plied. Therefore, the amount of diesel fuel transported to the
electrical and heating equipment” of the Teuri and Yagishiri Islands
Yagishiri electric power plant will be reduced, and the environ-
microgrid was optimized using a genetic algorithm (GA). In addi-
mental impact will decrease.
tion, the installation location of the renewable energy equipment
and the renewable energy utilization rate and consumption char-
acteristics of a fossil fuel were investigated. The objective of this 2.2. Teuri and Yagishiri Islands interconnection microgrid system
study is to propose a method to optimize the layout planning for
the renewable energy equipment introduced into an isolated island Fig. 2 shows the configuration of the Teuri and Yagishiri Islands
microgrid. The suitable installation location of renewable energy is interconnection microgrid system assumed in this paper. The de-
expected via numerical analysis. The design method for the opti- tails of the system configuration in Fig. 2 include the output of the
mization of the operation method of the microgrid based on the photovoltaic and wind power generators of Teuri Island, the heat
above output power is the goal of this study. pump load of both islands, and the electrical power load inter-
connected by the transmission lines. The power fluctuations in the
2. System transmission network are controlled by the control governor of the
backup power supply (diesel generator), as shown in Fig. 3. A
2.1. Teuri and Yagishiri Islands transmission network (including the distribution of electrical en-
ergy) and a heating (hot water) network are installed on each is-
As shown in Fig. 1, Teuri and Yagishiri Islands are isolated islands land; the transmission networks of both islands are interconnected
in northwestern Hokkaido, which is a cold, snowy area with a total by a submarine cable, which passes through each substation. The
population of 583 (2014) and an annual tourist population of backup power supply is connected to the transmission network of
18,000. The mean annual air temperature on the islands is 7.7
C Yagishiri Island. On the other hand, because the generator is not
(minimum: 8.9
C; maximum: 24.6
C), and the mean annual installed on Teuri Island, the photovoltaic power station and wind
precipitation is 1282 mm. The electrical power consumed by Teuri power generators are connected to the transmission network. The
and Yagishiri Islands is supplied by a diesel generator with 1110 kW heating network is connected to heat pumps, with a heat storage
rated power, which is located on Yagishiri Island, about 10 km from tank installed in each residence. The heating network assumes
the Yagishiri electrical power plant. The Teuri substation is con- small-scale hot water piping.
nected by a submarine cable. A small-scale solar cell is currently
installed on the rooftops of the Teuri elementary and junior high 2.3. Subject of isolated island microgrid
schools. However, there is no backup power supply that can meet
the power consumption needs of the entire island in case the The problem with an isolated island microgrid is that the wind
electrical power supply stops because of a problem with the environment changes significantly with geographical features.
Yagishiri electrical power plant or submarine cable. The total Therefore, it is expected that the introduction of wind power
annual power consumption for both islands is 2738 GW: the heat- generation to an isolated island has large differences in output
ing value for space heating, hot water supply, and baths is 18030 GJ; characteristics by installation location. Moreover, because the
the heating value of the propane required for cooking is 426 GJ; and conversion efficiency of a solar cell is dependent on battery tem-
the total annual energy consumed is 21200 GJ. Each year, 630 m3 of perature, the wind environment influences the production of
diesel fuel for generators (A-type heavy fuel oil), 10355 m3 of pro- electricity by photovoltaics. Consequently, the optimal arrange-
pane, and 546 m3 of kerosene are transported to both islands. ment of renewable energy equipment in consideration of the wind
 S. Obara et al. / Energy 161 (2018) 1211e1225 1213

Fig. 2. Teuri and Yagishiri Island interconnection microgrid.

Fig. 3. Energy model of the Teuri and Yagishiri Island interconnection microgrid.

environment needs to be included in the design of an isolated is- heat transfer relationship between the solar cell and the sur-
land microgrid. In this paper, a 3D topographical map and renew- rounding environment using the heat-fluid solver, as shown in
able energy equipment are modeled as computer-aided design Fig. 6. The temperature of the semiconductor layer inside the solar
(CAD) data. A heat-fluid analysis solver is then applied, the optimal cell was determined by calculating the heat transfer amount. Here
locations of the photovoltaics and wind power generators are ob- the mean temperature of the semiconductor layer was considered
tained by the analysis result. Moreover, corresponding to the as the temperature of the solar cell.
output fluctuations of the fringe (from a few minutes to about Equation (1) is the heat balance equation of the solar cell, where
20 min) of renewable energy, the application of a heating time shift qgh , qca , and qtr are the internal heating value of the solar cell, the
of by using a heat pump and heat storage tank is effective. For a amount of heat transferred by the convective heat transfer of the
microgrid with a high heat-to-power ratio, a large amount of wind, and the amount of radiant heat, respectively. In addition, Gpm
introduced renewable energy can be expected with a heat change and cp;pm are the mass and specific heat of the solar cell,
in fringe electrical power fluctuations and the application of a time
shift using a heat storage tank.

3. Introductory method of renewable energy

3.1. Photovoltaic power station

Fig. 4 shows the installation method for the array of the

photovoltaic power station. The solar cells are assumed to be
single-crystal modules, and their structure is shown in Fig. 5.
Because the conversion efficiency of a solar cell depends on its
temperature, the semiconductor temperature inside a solar cell was
analyzed by calculating the heat transfer of the structure shown in
the figure. The temperature of the solar cell was obtained from the Fig. 4. Large-scale photovoltaics.
1214 S. Obara et al. / Energy 161 (2018) 1211e1225

qpm;t ¼ h$qsr;t ; (4)

where h is the conversion efficiency of the cell. Because the tem-

perature of the semiconductor layer shown in Fig. 6 influences the
conversion efficiency, the conversion efficiency h was obtained
using Tpm;t from Equation (1).

3.2. Location for the installation of renewable energy facilities

Fig. 7a shows the outline and main geographical features of

Teuri Island. The northwestern area is a cliff zone, whereas resi-
dential and public buildings are built along the northeast coastline
and in the southeast area. The central zone of the island is a grassy
Fig. 5. Solar cell model. plain. Fig. 8 shows the occurrence distribution of the average wind
direction of Teuri Island for the past ten years. Although west by
northwest to southwest winds are common on Teuri Island, wind
respectively. The temperature change ðTpm;t  Tpm;t1 Þ of the solar states that are different from those in Fig. 8 are expected depending
cell for the sampling time from t  1 to t was obtained from on the location on the island. Wind power generators were
Equation (1): installed at any of the locations WP1eWP5 in Fig. 7b, and the
photovoltaic power station was installed to any of the locations

 PV1ePV4. These locations were confirmed during a field survey.
qgh;t þ Gpm $cp;pm $ Tpm;t  Tpm;t1 ¼ qca;t þ qtr;t : (1) Therefore, we could identify the locations for installing the wind
power generators and photovoltaic power station, where the
As shown in Equation (2), the production of electricity qpm of the
output would be stabilized at a maximum value.
solar cell consists of heat generation caused by internal resistance
qgh and the electrical power supplied qdc to the demand side.
3.3. Calculation of electricity production by wind power generators
and photovoltaic power station
2 2
qpm;t ¼ qdc;t þ qgh;t ¼ RLoad $Ipm þ Rir $Ipm (2)
First, the 3D topographical map shown in Fig. 7b was prepared
From the reflection factor gr and the amount of global solar using 3D CAD. The 3D CAD data of the wind power generators and
radiation qgr on the solar cell surface, the amount of insolation qsr , photovoltaic power station were placed at locations proposed for
which contributes to the power generated by the cell, is shown installation of the renewable energy equipment shown in Fig. 7.
below: Next, the temperature distribution at the surface of the solar
modules was analyzed using the heat-fluid analysis solver
described in Section 4.1. Regarding the streamline of the space
qsr;t ¼ ð1  gr Þ$qgr : (3)
surrounding the solar modules and wind power generators, the
Finally, the production of electricity qpm of the solar cell can be NavierStokes equation for laminar and turbulent fluid flows was
expressed as follows: used in the solver, and the turbulent flow kinetic energy was

Fig. 6. Model of charge and discharge of electricity of photovoltaics.

 S. Obara et al. / Energy 161 (2018) 1211e1225 1215

power generator, the power generation output was obtained. In

addition, the temperature of the semiconductor inside a solar cell
was found through heat transfer analysis, based on the speed and
direction of the wind flowing into the photovoltaic power station,
as obtained by the heat-fluid analysis solver. The average conver-
sion efficiency of the photovoltaic power station was obtained from
the results of the solar cell temperature and amount of insolation
on the solar cell shown in Fig. 6. Therefore, the electricity produc-
tion on arbitrary dates was estimated for the locations of the wind
power generators and photovoltaic power station indicated in
Fig. 7b.

4. Analysis method

4.1. Heat-fluid solver

The wind state depends on the geographical features around the

islands, the wind conditions at the wind power generator and
photovoltaic power station sites, and the electrical power output of
each power generator. Therefore, a heat-fluid analysis was per-
formed to estimate the inlet wind velocity toward the wind power
generators and the solar cell temperature (semiconductor tem-
perature) on Teuri Island. Based on the results, the output of the
wind power generators and photovoltaic power station was
calculated, and the optimal installation location of each power
generator on Teuri Island was determined. The geographical fea-
tures of Teuri Island, the shape of each solar cell, and other pa-
rameters were modeled from 3D data using 3D CAD software
(SolidWorks R2013). The heat flow was analyzed using a heat-fluid
solver (Flow Simulation 2013), taking into consideration factors
Fig. 7. Teuri Island: (a) Geographical feature of Teuri Island; (b) Proposed sites for wind
such as the wind state around the island, outside temperature, and
power generators and photovoltaic power plant. amount of insolation. The power outputs were estimated based on
the inlet wind velocity to the wind power generators and the
temperature of the semiconductor inside a solar cell (described
analyzed through the transport equation [13]. When the speed of below as the solar cell temperature). The optimal installation lo-
the wind flowing into the turbines, obtained by the heat-fluid cations of the wind power generators and photovoltaic power
analysis solver, was introduced into the power curve of the wind station on Teuri Island became clear from the results of the power
outputs. In Flow Simulation 2013, meteorological data (outside air
temperature, slope solar radiation, wind direction, and wind
speed), solar position (property values regarding the reflection and
dispersion of solar radiation with the ground and structures), and
the heat physical properties of the structure were given to the
analysis solver. The past meteorological data of Japan [14] were
obtained from the Meteorological Agency [15], the Japan Weather
Association, and the New Energy and Industrial Technology
Development Organization [16].

4.2. Energy balance equations

When the planned installation locations of the photovoltaic

power station and wind power generators were determined, the
operating cost (sum of fuel expenses, equipment costs, and facility
maintenance costs) for supplying the electricity and heat load to
both islands was investigated. Moreover, an energy system with
sufficient economic efficiency was proposed by optimizing the
installed capacity of the energy equipment for both islands. Equa-
tions (5) and (6) are the electrical power balance and heat balance
of the proposal Teuri and Yagishiri Islands' interconnection
microgrid (shown in Fig. 3), and Equation (7) is an expression of the
heat storage relations. Therefore, in Eq. (5), the sum of the elec-
tricity output from the enginn generator Eeg;t , the output of pho-
tovoltaics Epv;t , and the wind-power output Ewp;t in sampling time t
Fig. 8. Annual wind direction distribution of Teuri Island. are equal to the sum of the power demand DEl;t , the heat-pump D
1216 S. Obara et al. / Energy 161 (2018) 1211e1225

Fig. 9. Energy demand on Teuri and Yagishiri Islands microgrid.

Ehp;t and the electricity loss DEloss;t . On the other hand, Eq (6) is the February), and the maximum heat-to-power ratio (20) was ob-
heat balance, the sum of the heat output tained for February.
from the enginn generator Heg;t , the output of heat pump Hhp;t , and
output of the heat storage Hst;out;t in sampling time t are equal to 4.3.2. Meteorological conditions
the sum of the heat demand DHl;t , the heat storage DHst;in;t and the Fig. 10 shows some meteorological data (wind direction, outside
heat loss DHloss;t . The heat storage Sst;t in sampling time t is ob- temperature, maximum wind speed, mean wind speed, and pre-
tained by adding heat storage Sst;t1 in sampling time t  1 to heat cipitation) for 2013, obtained from the observatory on Yagishiri
input and heat output of the heat storage tank (this value is subtract Island. Because Yagishiri and Teuri Islands are very close, the
Hst;out;t from DHst;in;t ). meteorological data of the Yagishiri observatory is useful for the
weather analysis of Teuri Island. The wind speeds were low from
Eeg;t þ Epv;t þ Ewp;t ¼ DEl;t þ DEhp;t þ DEloss;t (5) June to August, and there was little precipitation from November to
March. Because Teuri and Yagishiri Islands are cold, snowy areas,
Heg;t þ Hhp;t þ Hst;out;t ¼ DHl;t þ DHst;in;t þ DHloss;t (6) there is a high demand for heating in winter (from November to
March), and because the wind speed is high in the winter, with an

 increased energy demand and little precipitation on the island, as
Sst;t ¼ Sst;t1 þ DHst;in;t  Hst;out;t (7)
shown in Fig. 10, using renewable energy is advantageous.

4.3.3. Wind power generators and photovoltaic power plant

Fig. 11 shows the size and power curve of the wind power
4.3. Analysis conditions generators used for the present analysis. Each wind power gener-
ator has a downwind-type horizontal axis propeller, a direction
4.3.1. Demand pattern of electrical power and heat control (active and passive yaw control), a security apparatus (pitch
Fig. 9 shows an estimation of the demand pattern of electrical control and centrifugal pitch control), and brakes (drum brake for
power and heat for Teuri and Yagishiri Islands on a representative parking and electromagnetic brake). The generator is inter-
day in each month. Although the demand pattern of electrical po- connected to a three-phase, 200 V electrical power network. The
wer consumed by household appliances, electric lamps, etc., was rated power of each wind power generator is 10 kW for a wind
almost the same each month, the heat load including space heating speed of 12.5 m/s; the cut-in sand cut-out speeds are 2.5 and
and hot water supply was larger in the winter (November and 20.0 m/s, respectively.

Fig. 10. Meteorological data for 2013.

 S. Obara et al. / Energy 161 (2018) 1211e1225 1217

Fig. 13. Performance of diesel power generator.

Fig. 14. Performance of heat pump system.

Fig. 11. Wind turbine generator. photovoltaic power station and wind power generators are 95% and
90%, respectively. The unit price of the facilities, maintenance cost,
and diesel fuel expenses used for the planned equipment capacity
The installation method of the array of the photovoltaic power are listed in Table 1. Each unit price of Table 1 is standard price in
station is shown in Fig. 4, and the relationship between the tem- Japan.
perature and conversion efficiency of each array is shown in Fig. 12.
4.3.5. Objective function and analysis flow
4.3.4. Backup power supply and heat pump Equation (8) is the objective function of the proposed system.
The output characteristics of the diesel generator (1110 kW rated The optimal operation of the system is the operating method of
power) currently installed on Yagishiri Island are presented in each piece of equipment for each sampling time in the case in
Fig. 13. Fig. 13 assumes an actual diesel generator (Homepage of which the value of the objective function is the minimum, after
Denyo Co., Ltd. http://www.denyo.co.jp/english/index.html;2012). completing each energy balance equation. When the fuel
In addition, Fig. 14 shows the performance coefficient using an air consumed by the diesel generator is minimized, the system
source heat pump [17]. The power conditioner efficiency of the configuration and the operation method are obtained as the
optimal solution.

X
12 X
23
M¼ Feg;m;t (8)
m¼1 t¼0

Table 1
Setting of cost.

Photovoltaics 325000 JPY/kW

Maintenance 5200 JPY/(kW year)
Wind power generation 210000 JPY/kW
Maintenance 4200 JPY/(kW year)
Diesel power generation 190000 JPY/kW
Maintenance 5510 JPY/(kW year)
Heat pump 100000 JPY/kW
Storage tank 21000 JPY/kWh
Power conditioner (photovoltaics) 75400 JPY/kW
Power conditioner (wind power generation) 100000 JPY/kW
Diesel fuel 92.2 JPY/L
Fig. 12. Conversion efficiency of solar cell.
1218 S. Obara et al. / Energy 161 (2018) 1211e1225

Fig. 15. Flow analysis of optimized operation.

 S. Obara et al. / Energy 161 (2018) 1211e1225 1219

Fig. 16. Analysis results of wind conditions for each site on Teuri Island: (a) Local wind speed and average wind speed; (b) Wind conditions for each proposed site.

Fig. 15 shows the analysis flow for the equipment planning and chosen times for wind speeds of 2.5e20.0 m/s. The wind direction
the operation method using the GA. The GA was developed using varies greatly, from west by northwest to east by southeast, and a
Microsoft Visual Cþþ 2013. The determination of each parameter of high-speed wind lands on the west side of the island. Fig. 16b shows
the GA yielded a range of suitable values through trial and error. the analysis results of the average annual wind speed and direction
Wind power generators are installed at WP5, a photovoltaic power at each position of the wind power generators shown in Fig. 7b. As
station is installed at PV3, and the heat balance [Equation (6)] and shown in Fig. 10, the wind direction on Teuri Island shifts frequently
power balance [Equation (5)] are calculated using the installed from west by northwest to east by southeast, and, as shown in
capacity set up for the trial. The installed capacities of the heat Fig. 16b, these winds flow into WP5 at high speed. The WP5 location
storage tank and heat pump are given so that a heat balance has the highest altitude on Teuri Island, and the wind blows from all
shortage does not appear for every sampling time on the repre- directions at high speed. Therefore, installation of the wind power
sentative day. Moreover, the total cost on the same day, including generators at WP5 is suitable. Fig. 17a shows the analysis results of
the equipment cost, operating cost (fuel expenses for the backup the velocity distribution for the case in which the wind on Teuri
power supply) for every sampling time, and facility maintenance Island blows from the north. Fig. 17b and c shows the analysis re-
cost, is calculated. In order to search for the minimum total cost, we sults for wind blowing from the southwest. Because of high altitude
changed the installed capacity and calculated the total cost again. in southwestern of the island, when the wind blows from the
The optimal installed capacity is that gave the lowest total cost. southwest, a slow-velocity backflow occurs in the northeast area.
Locations of isolated island change wind state largely and various
5. Results and discussion wind states have considerable influence on small-scale micro-grid
simultaneously. However, there are few examples concerning the
5.1. Installation location of wind power generators optimization of the installation location of wind power generators
based on investigation, as mentioned above, and the wind state of
Fig. 16a presents the annual wind direction on Teuri Island at the island.
1220 S. Obara et al. / Energy 161 (2018) 1211e1225

Fig. 17. Analysis results of a wind condition example.

5.2. Installation location of photovoltaic power station have released details by reference concerning comparison of
experimental results and analysis results [18].
As described in section 5.1, the wind state changes largely with
locations the isolated island. As a result, the surface temperature of 5.3. Equipment planning
photovoltaics is dependent on installation location, the annual
energy production of the solar cell has a difference. Example of Table 2 lists the output rates of the photovoltaic power station
investigations of an island microgrid in consideration of the wind (PV3) and wind power generators (WP5) introduced into the sys-
state is not known in the past. Fig. 18 (a) shows the wind speed tem used for analysis. The rated power of the photovoltaic power
distribution on Teuri Island for southwest wind, and Fig. 18 (b) station and wind power generators was not set to the system
shows the analysis results for the temperature distribution of the configurations (systems AeE) in Table 2. Instead, it is assumed that
solar cell semiconductor layer at the time of installing a photovol- the rates in Table 2 are based on the results of the wind state
taic power station at PV2. From the results of Fig. 18 (b), the mean analysis. Normally, the electrical power and heating demands on
temperature of the semiconductor layer differs by 10
C at most. both islands are met by diesel and kerosene power generation.
The conversion efficiency of the solar cell is shown in Fig. 19 based Fig. 20 shows the optimal system configuration, based on the
on the results of Fig. 18 (b) relative to the solar cell-temperature analysis results (Table 2), for the minimum cost, including the fa-
conversion efficiency (Fig. 12). Authors have described the details cility and maintenance costs, for an operation period of 20 years. It
of analytic accuracy in literature of [18]. Fig. 19 (a) shows an also shows a conventional system for operation periods of 15 and
example of the analysis results of the solar cell temperature 20 years. The results for the conventional electrical power system
installed at PV1 to PV4 on a representative day for every month. are shown in Fig. 20, and the operation optimization of the system
When a solar cell is installed at PV4, because there is little cooling of described in Fig. 15 is not included in these results. Fig. 20 shows
the module by wind, the solar cell temperature rises during April the total cost of the Teuri and Yagishiri Islands interconnection
and July; consequently, as shown in Fig. 19 (b), for both months, the microgrid systems C and D. The total cost of these systems for the
conversion efficiency is low. Because the conversion efficiency at operation period of 20 years is the same as that incurred in 17 years
PV3 is a little higher than that at other installation locations of operation under the conventional system. The renewable energy
(Fig. 19), the photovoltaic power station is installed there. Authors power supplied for each system configuration during times of total
 S. Obara et al. / Energy 161 (2018) 1211e1225 1221

Fig. 18. Analysis results of solar cell skin temperature.

minimum cost is 10%, 20%, 40%, 40%, and 30% of that for the backup results of the operating method of the heat pump and heat storage
power supply of systems A, B, C, and D, respectively. The supply rate tank. The power consumption of the heat pump (Fig. 21f) is ob-
for systems C and D is high, and their total cost is the lowest. tained from the result of Fig. 22b. Because the method of dis-
Therefore, the installation location of the photovoltaics and wind charging surplus heat through a radiator is not included in this
power generators was planned appropriately: when the Teuri and analysis, and the amount of heat storage in the summer season is
Yagishiri Islands interconnection microgrid follows the equipment planned to be very large, as shown in Fig. 22d. Most of the heat
configurations of systems C and D, it is expected that the electrical storage amounts in the summer are surplus heat, and the actual
power supplied will reach nearly 40%. heating storage capacity can be considerably reduced.
Fig. 23 shows the analysis results for the optimal operation of
5.4. Optimal operation method the system including the output of the photovoltaics and wind
power generators installed at each location. The fuel consumption
Fig. 21 shows the results of the optimal operation electrical of the backup power supply, the installed capacity of photovoltaics
power solutions when photovoltaics and wind power generation and wind power generators when the renewable energy was
are installed at PV3 and WP5 with suitable conditions. Fig. 21 was installed at PV3 and WP5 was set at 100% in Fig. 23a. Moreover, the
obtained by the operation optimization analysis using the GA rate of renewable energy (sum total of photovoltaics and wind
described in Fig. 15. On the other hand, Fig. 22 shows the results of power generation) to the energy output of a system is also shown in
the optimal operation solution concerning heat. Figs. 21a and 22a Fig. 23a and b. As shown in Fig. 23a, when renewable energy is
show the pattern of the electrical power and heat loads in a installed at PV1eWP4 and PV4eWP4, although the installed ca-
representative day each month. Fig. 21c and d shows the output pacity of renewable energy decreases, the renewable energy rate
results of the photovoltaics and wind power generation, and falls significantly. On the other hand, Fig. 23b shows the analysis
Fig. 21e is planned for driving the backup power supply. The fuel results of the output rate of photovoltaics and wind power gener-
consumption of the backup power supply in Fig. 21b is obtained ation. When renewable energy is installed at PV3 and WP5, even
from the results described above. though the power generation rate of wind power generation in-
According to the operation of the backup power supply in creases nearly 6.3% compared with the installations at other loca-
Fig. 21e, the engine exhaust heat of Fig. 22c is obtained. In the case tions, the renewable energy rate increases from 8.1 to 10.5%
for which the heat load of Fig. 22a is not filled with engine exhaust compared with those locations. The planning of the installation
heat only, it is necessary to satisfy any deficiency by operating the location and system management with the optimal wind state
heat pump and heat storage tank. Fig. 22b and d shows the analysis conditions achieves the increase in the renewable energy
1222 S. Obara et al. / Energy 161 (2018) 1211e1225

Fig. 19. Analysis results of photovoltaic performance.

introduction rate. estimated values of the electrical power output by renewable

energy obtained from the above results, the method of
6. Conclusion optimizing the operation of the main electrical and heating
equipment was developed using a GA (genetic algorithm).
In this study, we investigated the economical equipment (2) The proposal analysis method was introduced aiming at a
required for the Teuri and Yagishiri Islands interconnection high rate of renewable energy outputs of an isolated island
microgrid system. Wind conditions affect the electricity production microgrid. As a result, an isolated island microgrid with high
by wind power generators. However, the wind state changes utilization of renewable energy (40% or more per year) was
greatly over these small and detached islands. Furthermore, the expected.
cooling effect of the wind controls the temperature rise of the solar (3) The search for a suitable installation location for renewable
cells, which affects their output. Therefore, it was necessary to energy and optimization of the operating plan for the system
determine a position where suitable wind conditions prevail for the clarified greatly increased the application rate of renewable
installation of the photovoltaic power station and wind power energy in an isolated island microgrid.
generators. Based on the present analysis, the following conclu-
sions can be drawn: The cases of an isolated island microgrid without a battery have
not yet been clarified, and a design method for ensuring a lot of
(1) The suitable installation locations of photovoltaics and wind introduction of renewable energy has yet to be investigated.
power generators were investigated from the wind state Therefore, this paper proposed a design example for an isolated
analysis using a 3D topographical map. Based on the island microgrid without a battery.
Because the electricity quality (voltage, frequency, waveform) of
Table 2
the transmission network, and the reliability and stability of an
Power rate of renewable energy of each system. electric power system are not known well in this study, a designer
needs evaluation of the electricity quality in the transient response
System configurations Photovoltaics Wind power
analysis of the electricity. Large fluctuation of renewable energy
System A 100% 0% influences the electricity quality of the transmission network.
System B 75% 25%
Therefore, an important future works is development of the plan-
System C 50% 50%
System D 25% 75% ned method of the system configuration with the control method of
System E 0% 100% electricity fluctuations.
 S. Obara et al. / Energy 161 (2018) 1211e1225 1223

Fig. 20. Calculation results of the cost evaluation when introducing the optimal renewable energy layout into the conventional electrical power system.

Fig. 21. Analysis results of electrical power equipment. Case of PV3 and WP5.
1224 S. Obara et al. / Energy 161 (2018) 1211e1225

Fig. 22. Analysis results of heat power equipment. Case of PV3 and WP5.

Fig. 23. Analysis results of each location: (a) Rate of fuel consumption and equipment capacity; (b) Rate of energy output.

Nomenclatures qgh Generation rate of heat of the solar module [kW]

qgr Global solar radiation [kW]
cp;pm Specific heat of the solar module [J/(g K)] qpm Generation power of photovoltaic [kW]
Eeg Electric power output of engine generator [kW] qrb Reflected solar radiation [kW]
Ehp Electric power load of heat pump [kW] qr Radiative reflection [kW]
El Electric power demand [kW] qsd Direct solar radiation [kW]
Eloss Electric power loss [kW] qsr The amount of insolation which contributes to power
Epv Electric power output of photovoltaics [kW] generation of a solar cell [kW]
Ewp Electric power output of wind power generator [kW] qtr Radiation quantity of heat [kW]
Eeg Electric power output of engine generator [kW] Rir Internal resistance [U]
Feg Fuel consumption of a backup power supply [kW] Sst The amount of heat storage [MJ]
Gpm Mass of the solar module [kg] t Sampling time [Hour]
Heg Heat output of engine generator [kW] Tpm Surface temperature of solar module [K]
Hhp Heat load of heat pump [kW] T∞ Outside temperature [K]
Hl Heat demand [kW] u∞ Velocity of air [m/s]
Hloss Heat loss [kW]
Hst Heat output of heat storage tank [kW]
Ipm Current of the solar cell [A]
Greek characters
M Objective function [MJ]
D Load
m Month
gr Reflectivity of solar radiation on the surface of the solar
qca Rate of heat of air current [kW]
module
qdc Generation power of direct current [kW]
h Conversion efficiency [%]
 S. Obara et al. / Energy 161 (2018) 1211e1225 1225

References [10] Mermelstein J, Cannova C, Cruz M, Anderson B. Field demonstration of a novel

reversible SOFC system for islanded microgrid energy storage. ECS Trans
2017;78:2907e12.
[1] Mohammad MH, Mohsen M, Gharaveisi AA. Optimum islanded microgrid
[11] Vasquez V, Ortega ML, Romero D, Ortega R, Carranza O, Rodríguez JJ. Com-
reconfiguration based on maximization of system load ability and minimi-
parison of methods for controllers design of single phase inverter operating in
zation of power losses. Int J Electr Power Energy Syst 2016;78:343e55.
island mode in a microgrid. Renew Sustain Energy Rev 2017;76:256e67.
[2] Neves D, Brito CM, Silva AC. Impact of solar and wind forecast uncertainties on
[12] Zhao Z, Yang P, Guerrero MJ, Xu Z, Green CT. Multiple-time-scales hierarchical
demand response of isolated microgrids. Renew Energy 2016;87:1003e15.
~o PSJ. A control plan for the frequency stability control strategy of medium-voltage isolated microgrid.
[3] Mehrasa M, Pouresmaeil E, Jørgensen NB, Catala
IEEE Trans Power Electron 2016;31:5974e91.
stable operation of microgrids during grid-connected and islanded modes.
[13] Technical reference of the SolidWorks flow simulation 2012, section 1. USA:
Elec Power Syst Res 2015;129:10e22.
Dassault Systemes SolidWorks Corporation; 2012. p. 3e14.
[4] Yamada T. Demonstration tests of microgrid systems for small remote islands.
[14] Japan Meteorological Agency. The past meteorological data. 2013. http://
J Inst Electr Installation Eng Japan 2011;31:907e11 [in Japanese)].
www.data.jma.go.jp/obd/stats/etrn/;2013.12.
[5] Saga Y. The power conditioner system for electric-power stabilization. Wind
[15] Tenki JP. Japan weather association. 2013. http://tenki.jp/past/?
Energy 2013;37:319e22 [in Japanese)].
month¼1&year¼2012.
[6] Kojima T, Fukuya Y. Microgrid system for isolated islands. Fuji Electr Rev
[16] New Energy and Industrial Technology Development Organization (NEDO).
2011;57:4 [in Japanese)].
Annual amount-of-insolation data base (METPV-11). 2013. http://www.nedo.
[7] Haboro-cho. Haboro eco-island concept demonstration project. 2017. http://
go.jp/library/nissharyou.html;2017.4.
www.town.haboro.lg.jp/gyousei/shisaku-plan/ecoislandproject.html.
[17] Mitsubishi Electric Co. The chilling unit for air-conditioning. 2017 [in Japa-
[8] Sachs J, Sawodny O. A two-stage model predictive control strategy for eco-
nese)], http://www.mitsubishielectric.co.jp/ldg/ja/products/air/lineup/
nomic diesel-PV-battery island microgrid operation in rural areas. IEEE Trans
central/airchiller/lineup_02.html.
Sustain Energy 2016;7:903e13.
[18] Obara S, Konno D, Utsugi Y, Morel J. Analysis of output power and capacity
[9] Zhao B, Yang Y, Zhang X, Xue M, Li P, Xu C, Zhou D. Implementation of a dual-
reduction in electrical storage facilities by peak shift control of PV system with
microgrid system with flexible configurations and hierarchical control in
bifacial modules. Appl Energy 2014;128:35e48.
China. Renew Sustain Energy Rev 2016;65:113e23.