Engineering, 2013, 5, 81-88
doi:10.4236/eng.2013.51b15 Published Online January 2013 (http://www.SciRP.org/journal/eng)
Copyright © 2013 SciRes. ENG
A Wind-Solar-Energy Storage System Leading to High
Renewable Penetration in the Island System of Kinmen
Yuan-Kang Wu1, Ching-Yin Lee2, Bo-Shiung Jan1, Yong-Qing Huang3
1Department of Electrical Engineering, National Chung-Cheng University
2Department of Electrical Engineering, Tungnan University
3Green Energy and Environment Research Laboratories, Industrial Technology Research Institute
Email:al len wu@ ccu.edu.tw, cylee@mail.tnu.edu.tw, itri960611@itri.org.tw
Received 2013
Abstract
Kinmen Island lacks fossil-fuel energy, However, it has rich potential for solar and wind energy resources because of its
excellent climate and geographical location. Therefore, a large scale utilization of the renewable energy sources is fa-
voured. In June 2004, the Kinmen Count y go vern me nt pub lis hed t he St rate gic Pla n for the Sust ainab le D evelo p ment o f
Kinmen, which focuses on maintaining the ecology of the islands. In the near future, a high penetration of renewable
power can be predicted to be installed in this island. Howeve r, a large renewable energy penetration into a diesel power
system would face technical and economic problems. Therefore, this study intends to discuss the system operation of
the Kinmen system and investigates the original unit commitment scheduling. Based on the simulation results, a new
unit commitment sche duling will b e proposed i n this work.
Keywords: Kinmen; Re newable Energy; Unit C ommitment Schedulin g
1. Introduction
There are many instances where renewable energy
sources have been developed including: 100% RES
project for El Hierro, Spain; Crete Island in Greece;
Gotland in Sweden; Rhode Island and the Hawaiian Isl-
ands of the United States; Australian King Island, and
the island country of Saint Lucia and so on. Additionally,
there are more than 40 islands a round the world that have
either proceeded with or plan to carry out the construc-
tion of independent renewable energy sources. Kinmen
Island, Taiwan, has also obtained support and subsidies
from the central government and has installed two wind
turb ine ge ne rator systems [1].
Located on the southeast coast of mainland China,
Kin me n Island lacks many natural resources. Therefore,
this island is deeply affected by the natural ecosystem
and tourism. With the economy moving from military
rule to touris m, the island has seen a lot of new co nstr uc-
tion. Tashan power plant in Kinmen burns an average of
42,000,000 l of fuel oil annually. However, its annual
power generation capacity was 250,000,000 kWh and
this required 50,000,000 l of fuel oil. Meanwhile, the
plant produces approx. 150,000-170,000 metric tons of
carbon dioxide yearly. The current cost of conventional
diesel generation in K inmen is NT$7.2 per kWh. This is
expected to rise to somewhere between NT $11 and
NT$14 by 2020. In order to make Kinmen more sustain-
able, it focused on issues such as how to effectively use
Kinmen’s natural resources, sunlight and wind power ,
and how to develop renewable energy sources through a
distributed power supply system instead of the concen-
trated p ower o ne no w in use. Kinmen ho pe s it c a n pla y a
role in preventing global warming, take advantage of
new technolo gies and create sustainable developmen t.
Wind turbine technology has significantly improved
over the last 20 years, allo wing for t he rapid growth of t he
wind penetration. Large-scale wind power penetration im-
pacts the electricity supply industry in many aspects and
leads to fundamental changes in electric power systems.
Integration of wind power into power systems presents
chall enges to power system planners and operators [2-3].
These challenges stem primarily from the stochastic nature
o f wi nd ; power syste ms have to i ncorpo rate fo r first time a
source of high uncertainty, high volatility, and low predic-
tability. The expected large wind energy penetration im-
pacts the electricity supply industry both technically and
economically. Energy storage is a viable solution to sup-
press the fluctuation of wind power [4-5]. It can act as a
buffer that isolates the rest of the grid from these frequent
and rapid power changes caused by renewable resources.
In this study, a simulation analysis will be carried out by
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
82
adding energy storage systems to the hi gh wind penetration
system in Kinmen.
The interconnection of large amount of renewable gen-
eration resources would result in a lower amount of inertia
on the electric system and a potential frequency control
problem [6-8]. That is, if more synchronous machines are
displaced by wind generation, the system inertia will de-
crease, making the power s ystem more sensitive to genera-
tion-load imbalances. Generally, the small size and lack of
external support of isolated power systems could result in
severe voltage dips due to disturbances and frequency sta-
bility issues [9-10]. The weaker is the system the larger and
are longer voltage dips. Additionally, in isolated systems
with high renewable energy penetration, they have low
inertia constant and insufficient primary frequency regula-
tion; therefore, frequency stability issue should be taken
care.
In this work, the development strategies of renewable
ener gy sources and future renewable energy planning in
Kinmen will be introduc ed. Additionally, system impacts
of large-scale integration of renewable energy in Kinmen
and corresponding system simulation analyses are imple-
mented in this study. Finally, a revised unit commitment
scheduling in Kinmen system is proposed.
2. The Devel opment St rategies of Renewab le
Energy Sources in Kiemen
2.1. Current Electricity Power System of Kin-
men Island
The power system in Kinmen area is a typical islanding
system with the highest voltage rating at 22.8kV. Kin-
men Power Company constructed a new power plant at
Tashan and also expanded the electricity capacity of
Xiaxing power plant already in use. Therefore, today,
Kinmen Island has two po wer plants. It can be divided
into Tashan and Xiaxing thermal power plants, Jins a
wind parks, Tashan, Juguang, Qushan, and Xiaxing
substations. Tashan thermal power plant has phase 1 and
phase 2 diesel generator units, including 4 diesel genera-
tor units for phase 1 with an installed capacity of 7.91
MVA for each unit and 4 diesel generator units for phase
2 with an installed capacity of 8.25 MVA for each unit.
Therefore, the total installed capacity of thermal power
plants is 84.94 MVA. These 8 diesel generator units are
grid connected through the 13.2kV/22.8kV step-up
transformers. Ea c h gener a to r unit has t wo co ntr o l mod es :
droop and isochronous controls. For the two Type-C
wind power generator units in Jinsha wind park, each
unit has a rated output of 2MW. After the step-up trans-
former to convert the voltage to 11.4KV, each set of four
units is connected to the bus of Qushan substation. The
single line diagram of the whole Kinmen system is
sho wn in Fig.1.
Figure 1 Single line diagram of the Ki nme n system
2.2. Survey and Analysis regarding the Potential
for Renewable Energy Sources
The climatic statistic data for Kinmen region during
2004 to 2012 is shown in Table 1. In recent years, sun-
light duration in Kinmen has averaged 154.77 h per
mont h. This i s hi gher tha n p a s t r ec o rd s whic h ind ic ate a n
average of 138.8 h per month. In fact, the monthly aver-
age sunlight duration in July and August is over 200 h
and can sometimes reach up to 251.49 h. I n Kinme n Ju l y
and August have the longest sunlight duration and also
have the highest temperatures. Thus, it might seem that
Kinmen has optimum conditions for the development of
solar energy, whether that be solar PV systems or solar
thermal water heater systems.
The wind fro m June to August in Kinmen is from the
southwest while in winter months it blows from the
northeast. In recent years, the maximum 10 min wind
speed has ranged from 8.29 m/s to 12.24 m/s. The max-
imum instantaneous wind speed has ranged from 16.8
m/s to 31.9 m/s. Most winds blow from the ENE, NE,
and N directio ns. Based o n the statistical d ata fr om Table
1, Kinmen Island is also suitable for the development of
wind power generation systems to replace concentrated
po wer plants.
2.3. Current Protection Scheme in Kinmen
Power System
The under-frequency load shedding scheme is currently
used in Kin men to prevent incidents fr om caus i ng system
crashes and blackout; its detail is s ho wn in T able 2. Tai-
wan Power Company (TPC) uses automatic under-fre -
quency load shedding relays on the feeders at Tashan and
Xiaxing substations as the main power defense strategy
for Kinme n’s power system. The sheddi ng op
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
83
Table 1. C limatic statis t ic data for Ki nmen region d uring
2004-2012.
Month
Max. 10min
Wind speed
(m/s)
Max 10 min Wind
directio n Sunlight
duration (hr)
Jan.
8.77
58.89
104.04
Feb.
9.13
67.78
97.93
Mar.
9.57
91.11
105.94
Apr.
10.46
108.89
107.37
May 9.87 134.44 134.31
Jun.
11.10
227.78
150.31
Jul.
10.36
226.67
251.49
Aug.
12.24
250
219
Sep.
11.34
143.75
186.81
Oct.
10.16
123.75
193.55
Nov.
9.2 3
95
152.98
Dec.
8.2 9
58.75
153.51
Avg.
10.04
132.23
154.77
Table 2. Load Shedding Protection Scheme in Kinmen .
Stage Frequency
(Hz)
Shedding
Load
(kW)
Load Sh ed din g feeders
Sat ge I 57.3 3951 Brewery I, Gugang,
Guning, Sayang
Sa t g e II 57 4083 B rewery II, B rewery III,
Hightec, D o ngKung,
Gingang
Sa t g e II I 56.5 3997 MinSang , Mi nchu, J i n-
shan, J i ndong , T ao c i
Satge VI 56 5996
Jinmengcheng, chengxi,
Minchuang, Wuder,
Taiwu, Huanghai, Xian-
gyang
eration is di vided into fou r sta ges: The first st a ge i s se t at
57.3Hz; the second stage is set at 57.0Hz; the third stage
is set at 56.5 Hz, and the fourth sta ge is se t at 56Hz. The
operating time of each relay and circuit breaker is 3
cycles and 45ms respectively.
2.4. Current Unit Commitment Rule in Kinmen
Power System
Diesel generators in an isolated-island system plays a
significant role. The diesel units can follow the load
variations by means of their speed/power control
mechanism. T he power that the diesel plant must supply
at a certain time equals the load demand minus the
available generation of renewable sources and energy
storage syste ms. Also , if a die sel unit operates at a small
fraction of its nominal power output, its efficiency is
considerably reduced. For these reasons, the diesel plant
output must never undershoot a certain value. Table 3
shows the current unit commitment scheduling in
Kinmen. If the net load (minus wind power) is less than
16MW, then only four units are ON and all other units
are OFF; in that situation, power penetration for each
diesel generator is 25% and system spinning reserve is
large than 12MW. For load in the interval
(16MW~17.6MW), the suitable unit scheduling is
obtained if 5 diesel units are ON. As the net load is
increased, the number of ON units is also increased.
However, this unit scheduling in Table 3 is designed
under low wind power penetration. Once the wind power
penetration is increased, this unit scheduling has to be
revised.
Table 3. Current Unit Commitment Scheduling in Kinmen.
Net System Load
(MW)
(subtract wind power
generation)
Ratio
of each
diesel generator
(%)
Spinning Reserve
(MW)
Ratio of Spinning
Reserve
(%)
Generation
(MW)
No
Generation
(MW)
No
<16
2
0
4
4
25 >12 >75
16 ~ 17.6
2
1
3.9
4
22.2
>13 .2
>75
17.6 ~ 20.6
2.6
1
4.5
4
21.8 >10.8 >52.4
20.6 ~ 24.0
2.6
2
4.7
4
19.6
>10 .5
>43 .8
24.0 ~ 31.9
2.2
2
5.5
5
17.2 >11.4 >35.7
31.9 ~ 43.6
2.6
2
6.4
6
14.7
>8 .6
>19 .7
43.6 ~ 55.4
2.6
3
6.8
7
12.3 >8.1 >14 .6
55.4 ~ 62.2 2.6 3 6.8 8 10.9 > 9.3 15
2.5. Future Renewable Energy Planning
in Kinmen
The Industry Technology Research Institute (ITRI) has
implemented wind and solar energy potential in Kinmen
region, which considers available public land and area,
climate statistical data (wind speed and solar radiation),
and other important factors. In addition to the current 2
wind turbines at Jinsa wind park, other two regions are
also suitable for wind turbine installation: one is located
at Housa area and the other is at the area between T ianpu
and Hukotun (the East of Kinmen Island); in the former
area, five 850kW wind turbines would be installed. In the
latter area, there is a potential for building 20MW wind
generation capacity. In addition to wind energy resource,
there is also a large amount of solar generation capacity
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
84
that can be installed in Kinmen Island. Three locations
have been planned to install PV panels, including Kin-
men airport (272kW), tourist service center (865kW),
and Kinmen university (86kW). If all of the planning
renewable energy capacity is installed in Kinmen Island,
the installed capacity of wind and solar generation is
28.25MW and 1.223MW respectively. Furthermore,
energy storage system with 2.5MW capacity is also
planned to install in this syste m.
3. System Impacts of Large-Scale
Integration of Renewable Energy
3.1. System Inertia and Spinning Reserve
Real time system inertia and spinning reserve are the
main parameters having influence on operation limits.
In particular, the issue of inertia is important for high
wind penetration in synchronized systems. System in-
ertia can be defined as the total amount of kinetic
energy stored in all spinning turbines and rotors in the
sys tem. W hen a n unbalance occurs bet ween generation
and demand, the inertia would limit the rate of change
of frequency. For a given network condition replacing
the diesel generation by wind power generation would
lead to a reduction in global inertia and spinning re-
serve. T his tr anslate s direc tly into lower operatio n li m-
its to cope with the frequency stability criterion.
Therefore, system operators should review the system
inertia performance to adjust the frequency control
scheme on seasonal or annual bases. As wind power
penetration increases, the system inertia information
becomes more important for system operators.
Sudden loss of supply or demand will result in fre-
que ncy de via tio n fro m the nomi nal freq uenc y. T he rat e
of change in frequency due to imbalance depends on
the system inertia. System inertia is directly propor-
tional to synchronously rotating mass in the system.
The general equation for calculating rate of change of
frequency using system inertia constant (H) is illu-
strated i n (1 ) b e lo w:
0
22
df PD
ff
dt HH
=⋅ +⋅∆
(1)
where H is the system inertia consta nt on system base;
D is the power system load d ampening valu e; f0 is the
frequency at the time of disturbance; ΔP = (PL-PG)/PG;
PL indic ates the load prior to generation loss; PG in-
dicate s system generation after loss; Δf is the change i n
frequenc y. Assuming load dampening D to be zero, (1)
res ul ts in a simplified equation as belo w:
0
2
df P
ff
dt H
∆= =⋅
(2)
Therefore, the inertia constant H and frequency
change
f presents an inverse proportion; that is, as
system disturbance occurs, a larger H value would re-
sult in smaller freque ncy c ha n ge .
3.2. Limit Criterions of Wind Power Penetration
In addition to syste m inertia, the number of on-line die sel
generator units and the corresponding spinning reserve
margin may also affect the lowest system transient fre-
quency. The more diesel generator units running, the
larger the corresponding spinning reserve margin is, and
the le s s t he d r o p o f the lowes t s yste m t ra ns ie nt fre q ue ncy
would be. In other words, to prevent low transient fre-
quency from triggering an under-frequency load shed-
ding relay as a principle during an incident, the number
of running diesel generator units should be increased
appropriately. However, the issue of low power genera-
tion efficiency and accelerated depreciation when power
output of each single generator unit is too low should
also be considered. Therefore, a set of appropriate oper-
ating modes for diesel generators to meet the system re-
liability and economic is important. Generally, the de-
termination of the maxi mum wind power penetration will
be determined by the following limit criterio ns: minimal
and maximu m power production criterion of the co nve n-
tional diesel plants, ramp rate of the diesel plants, dy-
namic penetration limit for transient frequency and vol-
tage stability, power quality impact, and protection
schemes and load management.
In the c ase of small i sland s ystems, t he sudd en loss o f
all available wind power is quite probable. This may
happen due to generator trips, grid faults, and voltage
sags that exceed the fault ride through capability of the
wind turbine ge ne r ato r s, and even due to fast increases of
the wind speed that exceeds the wind turbine cut out
speed. Tripping wind turbines or diesel u nits will i mpose
a substantial frequency excursion in the Kinmen system.
Therefore, the transient frequency stability becomes a
significant crite rion to evaluate the maxi mum wind po w-
er penetration.
3.3. Case Study - the Impact of System Incidents
Limit Criterions of Wind Power
Penetration
In Kinmen ’s case, this dispatch and output arrangement
was chosen by mainly considering the tolerance of the
reserve margin of the system and the minimum accepta-
ble output o f the generator u nits. In the tran sient simula-
tion, six systematic perturbations were assumed sepa-
rately, including N-1 diesel generator units at Tashan
P ower Plant (the maximum output of generator units),
trip-off of the Jinsha wind farm, trip-off of the distribu-
tion line from Bus 2201 to Bus 2202, trip-off of the dis-
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
85
tribution line from Bus 2202 to Bus 2203, trip-off of P V
generators, trip -of f o f th e tr an sfo r mer a t Ju guan g Sub sta-
tion, and the occurrence of three-phase short-circuit
ground faults of important buses, so as to analyze the
effects o n the system freq uency and voltage.
Ta ble 4 sho ws the T r ansie nt Fr eque ncy and Vol tage i n
Kinmen Power System under various incidents. In this
study, system load is assumed as 20MW and the minimal
power production of the conventional diesel plants is
assumed to be 50% install capacity. Table 5 concludes
the simulation results at the similar operation conditions
but the system load is increased to 50MW. We can find
from Tables 4 and 5 that the transient stability impact is
limited if distribution lines, transformers, or PV genera-
tors trip offline. Ho wever, o nce one o f the diesel units at
Tashan or the Jinsha wind farm trips offline, then tran-
sient frequency dip increases, which would activate
low-frequency load shedding relay. Therefore, in this
study, we focus on the transient analyses based on the
trip-off of a diesel unit at Tashan Po wer Plant or the Ji n-
sha wind fa rm.
Table 4. T ransient Fr equency and Voltage in Ki nmen Power System under various incidents (System Load=20MW).
Type of System Incidents
Bus 2201 Bus 2203 Bus 1104 B us 1105
Frequency
(Hz)
Voltage
(V)
Frequency
(Hz)
Voltage
(V)
Frequency
(Hz)
Voltage
(V)
Frequen cy
(Hz)
Voltage
(V)
Li ne trip (from 2201 to 2202) max 60.01 1.003922 60.01 0.99 60.01 1.00 60.01 0.998
min 59.99 0.9973 64 59.99 0.99 59.99 0.99 59.99 0.991
Line trip (from 2202 to 2203)
max
60.01
1.00
60.01
0.99
60.01
1.00
60.01
0.99
min 60 0.99 60 0.98 60 0.99 60 0.98
PV generat or s trip
(bus 110 1)(0.865MW)
max 60.16 1.01 60.16 1.01 60.16 1.00 60.16 1.00
min 59.51 1.00 59. 51 0.99 59.51 1.00 59.51 0.99
Transformer trips
( Juguang Subst ation MTR2)
max 62.57 1.0 2 62.56 1.02 60 1.00 62.56 1.01
min 59.25 1.00 59.25 1.00 60 0 59.25 0.99
Tasha n Unit N-1 trips (4.125MW)
max
61.15
1.0 1
61.15
1.00
61.15
1.0 1
61.15
1.00
min 56.86 0.99 56.86 0.99 56.86 0.99 56.86 0.99
Jinsha wind farm trips (2.09MW) max 60.41 1.00 60.41 0.99 60.41 1.00 60.41 0.99
min 58.81 1.00 58.81 0.99 58.81 0.99 58.81 0.99
Table 5. Tr ansient F requency a nd Voltage in Kinmen Power System unde r var ious incidents (Sys t em Lo ad= 50MW).
Type of System Incidents
Bus2201 Bus 2203 Bus 1104 Bus 1105
Frequency
(Hz)
Voltage
(V)
Frequency
(Hz)
Voltage
(V)
Frequency
(Hz)
Voltage
(V)
Frequency
(Hz)
Voltage
(V)
Li ne trip (from 2201 to 2202) max 60.01 1.00 60. 01 0.99 60.01 0.99 60.01 0.99
min 59.99 1.00 59.99 0.99 59.99 0.99 59.99 0.99
Line trip (from 2202 to 2203) max 60.006 1.00365 60.01 0.99 60.01 1.00 60. 01 0.99
min 59.99 1.00 59.99 0.99 59.99 0.99 59.99 0.99
PV generat or s trip
(bus 110 1)(0.865MW)
max 60.07 1.00 60.07 0.99 60.07 1.00 60. 07 0.99
min 59.76 1.00 59.76 0.99 59.76 0.99 59.76 0.99
Tasha n Unit N-1 trips (4.125MW) max 60.44 1.00 60. 44 1.00 60.44 1.00 60.44 1.00
min 58.71 1.00 58.71 0.99 58.71 0.99 58.71 0.99
Jinsha wind farm trips (2.09MW) max 60.37 1.01 60.37 0.99 60.37 1.00 60.37 1.00
min 58.90 1.00 58.90 0.99 58.90 0.99 58.90 0.99
4. System Simulation Analyses
In this paper, the modeling and simulation study has
been carried with the PSS/E software. PSS/E is com-
posed of a comprehensive set of programs for studies of
power system transmission network and generation per-
formance in both steady-state and dynamic conditions.
Both steady-state and dynamic simulations have been
carried out to assess the power grid behavior of the Kin-
men Island system under study. The main goal of this
study is to analyze the impact of wind generation on the
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
86
stability of the grid. Two main aspects have been cov-
ered: the impact of wind generation on frequency stabil-
ity and on transient stability.
4.1. Load Flow Studies
It is important to undertake load flow studies in the de-
sign phase to ensure the voltage distribution over the
network will remain within statutory limits. In addition,
cables and transformers must be also dimensioned ade-
quately to ha nd le the thermal l o a di ngs. In thi s st udy, load
flow studies were performed for several scenarios in or-
der to simulate the range of anticipated load demands
and generation mixes on the island. According to load
flow analysis results, the load ratio of each line is in the
range of 16-77% after wind power is integrated into the
system.
4.2. System Transient Analyses
In this st udy, the off-pea k syste m in Kinmen i s con sidered,
including 18.2MW load, 4MW wind power output, and
0.528 PV output. This adopted system condition is under
the maximum renewable generation penetration (24.8%)
in Kinmen region currentl y. Based on the current unit
scheduling at the Tashan Power Plant, four diesel units
have to be operated at least.
Due to the low PV penetration (2.89%), this study
ignor es the e ffec t of PV and d iscusse s the s ystem i mpac t
from three-phase short circuit fault, trip-off of a diesel
unit, and trip-off of the Jinsha wind farm respectively.
Furt her, the o riginal unit co mmitment sc heduli ng will b e
revised based on the simulation results.
Case 1: Three-phase short-circuit grounding fault at Bus
2201 for 6 cycles
In thi s c ase , t he most severe symmetrical three-line-to -
ground fault is considered as a network disturbance; it i s
assumed that wind speed is constant and equivalent to
the rated speed for the wind tur bines. Figure 2 shows t he
dynamic response of one wind generator at Jinsha when
a three-phase short circuit fault occurs at 22.8kV Tashan
Bus 2201. The fault occurs at 1 sec and is cleared after
1.1sec. During the transient process, each wind turbine
will decrease its active power but support additional
reactive power to try to maintain its voltage output. The
reactive power of 0.974 MVAr is produced during the
fault for eac h wind turb ine. After the fault is cleared, the
wind t urbi ne re tur ns to supp ort stable a ctive po wer to the
grid. Figure 3 shows the system transient voltage curves
at bus 2201 during the event; it is noted that the voltage
drop is obvious and the recovery voltage returns to the
normal system voltage after the fault. In this case, the
system minimum frequency is 59.86Hz therefore the lo w
frequency relay wo uld not be activated.
Figure 2 Pow er output of a wind turbi ne at J ins ha (Case1)
Case 2: One of the diesel gene r a tors trips o ff for 6 cycles
It is assumed that one of the diesel generators at Tashan
Power Plant trips instantaneously to be offline, and the
voltage and frequency curves at bus 2201 are shown in
Figure 4. It is observed that the transient frequency is
decreased to 57.41Hz and there is an oscillation
phenomenon on recover y voltage and frequency curves.
Figure 3 Transient voltage and frequency curves at Bus
2201 in the Kinmen system under case1
Figure 4 Transient voltage and frequency curves at
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
87
Bus 2201 in the Kinmen system under case 2
Figure 5 shows the transient active power, reactive
power, and output voltage from a wind turbine at Jinsha
wind farm. It can be observed that the transient voltage
dip is small. Therefore, the impact of this fault case on
the transient voltage is limited.
Case 3: Jinsha Wind Farm trips o ffline
It is assumed that all wind turbines in the Jinsha wind
farm trip instantaneously to be offline, and the transient
voltage and frequency curves at bus 2201 are shown in
Figure 6. It is observed that the transient frequency is
decreased to 57.78 Hz and there is an oscillation
phenomenon on recovery voltage and frequency curves.
However , the frequency dip is les s than that in Case 2.
Figure 5 Power output of a wind turbine at J ins ha (C ase 2)
Figure 6 Transient voltage and frequency curves at Bus
2201 in the Kinmen system under case 3
5. Revised Unit Commitment Scheduling in
Kinmen System
In this study, the current install capacity of renewable
ener gy gene ra tio n i n K i n men r egi on i s cons id er ed ; that i s,
the total renewable generation is 4.528MW. Then system
transient analyses under the load between 18MW and
50MW are implemented. The analyzed system incidents
include the trip-off of Jinsha wind farm and the trip-off
of one diesel unit. Based on the simulation results, a
revised unit co mmitment scheduling is suggested.
In order to make sure the operation safe, the spinning
reserve has to maintain the maximum capacity of a diesel
generator. Additionally, it should be noticed that the
diesel generators should be avoided to operate under low
or medium load. Finally, the system operation should be
maintained not to activate the low frequency load
shedding relay. Table 6 shows the suggested new unit
commit ment scheduling based o n our simula t io n results.
6. Conclusions
The power system on the Kinmen Island is presented.
The renewable generation is sited at various locations
around the Kinmen network, and includes wind turbines
and inverter-connected PV systems. Several simulations
were performed in this research to study the impact of
the wind farm and PV generator integration on the dy-
namic beha vio r o f the Ki nme ns po wer syste m. T he con-
sidered grid disturbances are the trip-off of a diesel ge-
nerator and the wind farm. Simulations have shown that
deviations of the power system frequency and voltage
would be unacceptable under several system incidents.
Ho wever, it is still possible to operate the power system
of Kinmen with a high level of renewable penetration
maintaining a high level of se curity if adequate spinni ng
reserve and unit com mi tment scheduling are available.
Therefore, this study has proposed a new unit commit-
ment scheduling to cope with the high renewable energ y
penetration.
7. Acknowledgements
The autho rs gr ate full y ack no wled ge the fi nanc ial s upp or t
by Bureau of Energy, Ministry of Economic Affairs,
R.O.C. under the Project "Development of Key Control
Technologies fo r Distrib uted Ener gy System".
REFERENCES
[1] Hua-Yueh Liu and Sung-De Wu, “An assessment on the
planning and construction of an island renewable energy
system A case study of Kinmen Island,” Renewable
Energy, Vol. 35, No. 12, 2010, pp.2723-2731
[2] V. Akhmatov and P. B. Eriksen, A Large Wind Power
System in Almost Island OperationA Danish Case
Y.-K. WU ET AL.
Copyright © 2013 SciRes. ENG
88
Study,” IEEE Transactions on Power Systems, Vol. 22 ,
No.3, 2007, pp. 937-943
[3] Eel-Hwan Kim, Jae-Hong Kim, Se-Ho Kim, Jaeho Choi,
Lee, K.Y. and Ho-Chan Kim, “Impact Analysis of Wind
Farms in the Jeju Island Power System,” IEEE Systems
Journal, Vol.6, No.1, 2012, pp.134 – 139
[4] Han sang Lee, Byou ng Y oon Shi n, San gchul Han, Seyong
Jung, B yungj un Park and Gil soo J a ng, “Compensation for
the Po wer Fluctuation of the Large Scale Win d Farm Us-
ing Hybrid Energy Storage Applications,IEEE Trans-
actions on Applied Superconductivity, Vol. 22 , No. 3,
2012, pp.Page(s ): 5 701904
[5] Delille, G., Francois, B. and Malarange, G., “Dynamic
Frequency Control Support by Energy Storage to Reduce
the Impact of Wind and Solar Generation on Isolated
Power System's Inertia,IEEE Transactions on Sustain-
able Energy, Vol .3, No.4, 2 012, pp.931-939
[6] Margaris, I.D., Papathanassiou, S.A., Hatzi arg yriou, N. D.,
Hansen, A.D. and Sorensen, P, Frequency Control in
Autonomous Power Systems With High Wind Power
Penetration,IEEE Transactions on Sustainable Energy,
Vol. 3, No.2, 2012, pp. 189-199
[7] Moya, O.E., “A spinning reserve, load shedding, and
economic dispatch solution by bender's decomposition,
IEEE Transactions on Power Systems, Vol. 20 , No. 1,
2005, pp .384-388
[8] Kristo f De Vos, Andreas G. Petoussis, Johan Dri esen and
Ronnie Belmans, “Revision of reserve requirements fol-
lowing wind power integration in island power systems,”
Renewable Energy, Vol.50, 2013, pp.268-279
[9] Stavros A., Papathanassio u and Nikos G. Boulaxis,
Power limitations and energy yield evaluation for wind
farms operating in island systems,” Renewable Energy,
Vol.31, No.4, 2006, pp.457-479
[10] J.K. Kaldellis, K.A. Kavadias and A.E. Filios, A new
computational algorithm for the calculation of maximum
wind energy penetration in autonomous electrical genera-
tion systems,” Applied Energy, Vol. 86, No.7–8, 2009,
pp.1011-1023
Table 6. New Proposed Uni t Commitment Scheduling in Kinmen.
Net System Load (MW)
(subtract wind power
generation)
Xiaxing Power Plan t Tashan Power Plant
Ratio of each diesel
generator (%) Spinning Reserve
(MW) Ratio of Spinni ng Reserve
(%)
Generation
(MW)
No
Generation
(MW)
No
<=13 2 0 4 4 25.0 >12 >75 .0
13~14
2
1
3.9
4
22.2
>13.2
>75 .0
14~22 2.6 1 4.5 4 21.8 >10.8 >52 .4
22~26
2.6
2
4.7
4
19.6
>10.5
>43 .8
26~36
2.2
2
5.5
5
17.2
>11.4
>35 .7
36~47 2.6 2 6.4 6 14.7 >8.6 >19 .7
47~55
2.6
3
6.8
6
12.3
>8.1
>14 .6
55 ~
2.6
3
6.8
7
10.9
9.3
15.0