The document establishes a comprehensive evaluation index system to assess wind power utilization levels in wind farms in China. It considers factors like wind resources, wind turbine types, wind power output, curtailment levels, grid technology, and operational management. It applies an improved analytic hierarchy process and fuzzy comprehensive evaluation method to evaluate the Hami wind farm in Xinjiang province. The results can help identify issues affecting utilization and guide wind farm planning and grid integration to improve the industry.
2. capacity and the efficiency ignored [10–12].
The method of multiple indexes is used to analyze the utilization
level of wind power, because it is closely related with many factors. In
the literature [2,13], three levels of wind power utilization index system
is constructed, and the system includes basic indicators, the develop-
ment of scale indicators and the use of efficiency indicators. Among
them, the annual wind power generation accounted for the proportion
of power consumption and wind power installed capacity accounted for
the proportion of total net installed capacity highlight the scale of the
development of wind power in various regions of our country, and the
annual utilization hours index is a comprehensive embodiment of
accommodating wind power in the power system.
For the evaluation methods, domestic and foreign scholars have
proposed dozens, and several methods have been widely used, such as
principal component analysis method (PCAM), Delphi method, the
grey comprehensive evaluation method (GCEM), artificial neural net-
work method (ANNM), analytic hierarchy process (AHP) and fuzzy
comprehensive evaluation method (FCEM), and so on, the advantages
and disadvantages of these methods as shown in Table 1.
Therefore, in this paper, the wind farm operation comprehensive
evaluation model is established, combined the improved analytic
hierarchy process (IAHP) and fuzzy comprehensive evaluation method
(FCEM), the status of the wind energy farm operation is comprehen-
sively evaluated. The results can facilitate the operation and planning
of wind farms and promote the scientific development of the wind
power industry.
2. Wind power in China
China is a large country with excellent wind energy production that
reaches approximately 3.226 billion kW. With technological advance-
ment, the wind energy reserves of the country amount to 1 billion kW,
which is close to that of the USA. Hence, China is regarded as one of the
five major wind power producers in the world [14–16].
The distribution of the annual duration of wind speed above 3 m/s
in China is shown in Fig. 2 [1,2]. Wind energy resources depend on
wind energy density and the annual cumulative hours of wind energy.
The distribution of the average wind power density in areas with a
height of 70 m is shown in Fig. 3 [17].
Wind energy resources are greatly influenced by the terrain. In
China, wind energy is mainly distributed in the following areas [18–
20]:
(1) Southeast coast and its islands, which serve as the largest wind
energy resources;
(2) In ner Mongolia and Gansu in the north, which are major wind
energy resources;
(3) Heilongjiang and Eastern Jilin and the Liaodong Peninsula coast,
which also provide a considerable amount of wind energy;
(4) The Qinghai Tibet Plateau and the three northern regions of the
northern coastal area, which serve as a large wind energy sources;
(5) Yunnan, Guizhou, Sichuan, Gansu, Southern Shaanxi, Henan,
Western Hunan, the mountainous areas of Fujian, Guangdong,
Guangxi, and the Tarim Basin, which feature the smallest wind
area.
With China's recent focus on the development of nine 10 million
kW class wind power bases, nine wind power bases have been built in
Hami in Xinjiang, Jiuquan in Gansu, and in other areas, as shown in
Fig. 4. The development of these wind power bases follows the wind
resource distribution in China and the layout involving the “building of
a large base for a large power grid.” The nine large wind farms are
expected to reach an installed capacity of more than 79 GW in late
2015; this value should account for more than 75% of the total wind
power of the country [21,22].
At present, China's wind power industry is a global leader in the
field of wind power development. Since 2010, the total wind power
installed capacity of the country has been ranked first [2]. By the end of
2012, the wind power of China had reached 13.5 TWh, thereby making
wind power the third largest type of power supplied by the country,
with thermal power and hydropower topping the list [3,4].
By the end of 2014, the cumulative wind power installed capacity
had reached nearly 114.6 million kW, which accounted for 7% of the
total power installed capacity during that period. Moreover, the grid
connected capacity reached nearly 1 million kW, which corresponded
to the operation of nearly 7 million wind power units, i.e., more than
1500 wind farms. Fig. 5 shows the wind power development in China
from 2001 to 2014 [2–4].
According to the China Wind Energy Association statistics [23], the
most pressing problem of the Chinese wind power industry is wind
power curtailment. The details of the wind power curtailment of China
in 2014 are illustrated in Fig. 1. During the said period, the total wind
power curtailment reached 70 billion kWh, which corresponded to a
direct economic loss of nearly 40 billion yuan over the past five years.
As a result of the rejection of wind power turbines for use in the power
grid, the poor power supply structure, and the limited regulation
capacity of the power system, the output of wind power turbines is
restricted, and wind power equipment is even shut down. These
conditions increase the severity of the rationing of disposable wind
power. Table 2 presents the wind power curtailment in major Chinese
provinces in 2014 [24]. The rate of wind power curtailment reached
Fig. 1. The wind power curtailment of China from 2010 to 2014 [2,4,5].
(Source: CWEA, GWEC. CWP: Curtailed Wind Power, PCWP:Proportion of Curtailed
Wind Power, Economic loss: ¥109
yuan).
Table 1
Comparison of several common comprehensive evaluation methods.
Evaluation
methods
Advantages Disadvantages
PCAM Effective reduction of the
number of original variables by
using dimension reduction
ideas, to achieve the effect of
fast convergence speed.
When the principal
component factor is positive
and negative, the
comprehensive evaluation is
not clear.
Delphi Give full play to the role of
experts, benefit by mutual
discussion,high accuracy
The process is complex, and
the time is long.
GCEM This method is applicable to
the problem of accurate and
objective index.
Only judge the pros and cons,
do not reflect the absolute
level.
ANNM This method can deal with
nonlinear and non local large
scale complex systems, and has
strong adaptability and fault
tolerance.
Requires a large number of
samples and marginal
conditions
AHP Level of clarity and ease of
analysis
Evaluation of the object of the
factors can not be too much,
generally not more than 9
FCEM This method can solve the
problem of fuzziness and
uncertainty.
If the indicators do not have
mutual independence, it is
difficult to solve the
information related issues
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
462
3. more than 20% in Jilin, Heilongjiang, and Inner Mongolia (IM). Wind
power curtailment was also serious in Xinjiang, Liaoning, and Hebei,
with the rate exceeding 16%.
3. Characteristics of an operating wind farm in China
According to the statistics of the National Energy Administration
and China Wind Energy Association (2010–2014) [24,25], the char-
acteristics of an operating wind farm in China are as follows:
(1) The randomness of the wind power output is strong, and inter-
mittence is obvious. The curve of the daily power output of a wind
farm in Hami in Xinjiang is shown in Fig. 6. The power output of
the wind farm is obviously unstable, and the intermittence and
volatility are evident.
The power output of a wind farm is influenced by wind speed
and the type of wind power unit. Under different wind conditions,
the power output of wind turbines of the same type may vary, as
shown in Fig. 7.
A strong wind speed equates to a large power output. Under the
same wind conditions, wind power capacity of different type wind
turbine may also vary, as shown in Fig. 8. At 0.5 s, the power
output of all wind turbines fluctuates. Specifically, the power
output of direct-drive wind turbines shows the most obvious
fluctuation, that of doubly fed wind turbines fluctuates frequently,
and that of asynchronous wind turbines is relatively stable. After
0.5 s, the power output of direct-drive wind turbines becomes
stable and reaches its peak, that of doubly fed wind turbines shows
slight fluctuations, and that of asynchronous wind turbines shows
the least fluctuation.
(2) The construction of large-scale wind farms does not match existing
power systems, and such condition leads to serious wind power
curtailment. Wind farm construction is faster than grid planning
for thermal power and hydropower. In some instances, wind power
and other power sources are inadequate, and peak shaving is weak.
When the acceptance ability of a power grid is limited, the
dispatching department remotely controls the wind turbine output.
In such a case, a large amount of wind power is curtailed.
Examples of wind power turbine output in a full load condition
and in a curtailment case are shown in Figs. 9 and 10, respectively.
A high wind power curtailment clearly equates to low wind power
generating capacity and significantly low wind power utilization
rate.
In summary, a large-scale wind power-connected grid influences
the secure operation of a power system because of the randomness and
volatility of wind energy. Thus, wind farms in China should be
beneficial and sustainable to ensure the healthy development of the
wind power industry. Along with the establishment of friendly wind
farms, provincial power grids must also be operated to improve the
level of wind power utilization, reduce the influence of wind farm
operations on power systems, and promote a positive interaction
between grid networks and wind farms.
4. Index system for wind power utilization level
In order to reasonably evaluate the development of wind power in
china and fully grasp the level of wind power utilization, a diversified
Fig. 2. Distribution of annual duration of wind speed above 3 m/s in China [1,2].
(Source: CWEA, China Meteorological Station).
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
463
4. wind power utilization level index comparison system is constructed in
the literature [2]. The future wind power utilization level is forecasted
and analyzed by using the overall optimization model of power system.
The development level of China's wind power should be comprehensive
study from multiple aspects. We should focus on the coordinated
development of wind power and power system level and improve the
overall power system planning and operation level, which is the key to
realize wind power scientific and efficient development.
In the literature [2], the overall utilization level of wind power
industry in China is studied, and this paper focuses on the utilization
level of a certain wind farm. In addition to the object of study, the
research methods are also different, in literature [2] the future wind
power utilization level in China is forecasted and analyzed by using the
overall optimization model of power system, however, in this paper the
wind farm operation comprehensive evaluation model is established by
using the method IAHP and FCEM.
Wind power penetration is used in a global scale to indicate the
level of wind power farm utilization. The level of wind power farm
utilization is related not only to the size of wind resources and the
installed capacity but also to the power system. In the present work, we
consider the characteristics of wind power farm operations and
determine the factors that affect the level of wind power utilization.
These factors include wind resource characteristics, type of wind
turbine, wind power productivity, equipment operation, wind power
curtailment, and friendly wind farm conditions [26].
4.1. Wind resource characteristics
The power output of a wind turbine is written as
P
π
ρD v C=
8
.W P
2 3
(1)
The electric energy is written as
W PT
π
ρD v C T= =
8
,W P
2 3
(2)
where ρ is the air density, v is the input wind speed, DW is the diameter
of the wind wheel, CP is the power factor of the wind turbine, and T
denotes the effective wind speed hours.
As shown, the factors that affect conversion of energy from wind
electricity in wind resources are the wind speed, air density, and
effective wind speed hours.
4.2. Type of wind power turbine
At present, wind farms in China come in the form of squirrel cage
induction generators, doubly fed induction generators, and direct-drive
permanent magnet wind generators [9].
Given that wind turbines have several types, their features tend to
differ. Specifically, their abilities to capture wind energy, the quality of
their power output, and the conditions that stabilize the power grid
after the occurrence of a fault are not identical. Their effects also differ
even with the use of the same model produced by different manufac-
turers. The types and characteristics of wind turbines are presented in
Table 3.
4.3. Operation of wind power equipment
With the development of the manufacturing industry dedicated to
Fig. 3. Distribution of the average wind power density in China's land 70 m height [16].
(Source: CWEA, China Meteorological Station).
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
464
5. wind power equipment and the improvement of wind power technol-
ogy, the cost of wind power has been reduced, but the quality of
equipment during operations remains problematic. The rate of wind
power equipment utilization and the repair duration in cases of
equipment failure are used to measure the quality of a wind turbine.
The rate of wind turbine utilization can be expressed as
η
T A
T B
=
−
−
× 100%.
(3)
In Eq. (3), A is the downtime for equipment failure or routine
maintenance, B is the non-equipment-related downtime, and T denotes
the statistical hours.
Fig. 4. Nine wind power bases in China.
Fig. 5. The development of China's wind power from 2001 to 2014. (NIC: New Installed Capacity, CIC: Cumulative Installed Capacity, GR NIC: Growth NIC %, GR CIC: Growth CIC %).
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
465
6. 4.4. Wind power productivity
4.4.1. Wind power generation capacity
The wind power generation capacity refers to the sum of the power
output per wind power unit during typhoon periods, i.e.,
∑E E= ,
i
n
i
=1 (4)
where Ei is the power output of each wind turbine at the outlet and n is
the number of wind turbines in a wind farm.
4.4.2. Wind power installed capacity
The wind power installed capacity is written as
∑P P= ,
i
n
i
=1 (5)
where Pi is the capacity of each wind turbine and n is the number of
Table 2
Curtailed wind power of China's key provinces in 2014. (CIC: Cumulative Installed Capacity, NGC: New Grid Capacity, CGC: Cumulative Grid Capacity, GE: Grid Electricity, CWP:
Curtailed Wind Power, PCWP: Proportion of Curtailed Wind Power, PCWP=CWP/(CWP+GE)).
Source: The National Energy Administration (http://www.nea.gov.cn/) [24].
Rank Province CIC/GW NGC/MW CGC/MW GE/TWh CWP/TWh PCWP/%
1 Jilin 4.65 689.2 4256.6 9.87 5.37 25.24
2 Heilongjiang 5.53 1336.3 5109.7 15.57 5.14 24.82
3 IM 22.31 2309.8 20185.6 83.22 16.9 23.03
4 Xinjiang 9.67 248.9 9012.0 25.32 5.88 18.85
5 Liaoning 7.11 1091.5 6953.8 20.02 4.24 17.48
6 Hebei 9.87 1729.8 9056.6 39.22 7.89 16.75
7 Gansu 10.73 2541.3 9768.7 27 1.77 6.15
8 Shandong 8.26 1282.8 8005.1 26.17 0.7 2.61
9 Shanxi 5.86 1921.0 5673.6 17.15 0.33 1.89
10 Ningxia 6.14 2501.3 5893.2 15.13 0.16 1.06
Fig. 6. The curve of the daily power output of a wind farm in Hami in Xinjiang.
Fig. 7. The same type of wind turbine output power in different wind conditions.
Fig. 8. Under the same wind conditions, output of different type wind turbines.
Fig. 9. Wind power turbine output in a full load condition.
Fig. 10. Wind power turbine output in a curtailment case.
Table 3
Comparison of the types and characteristics of wind turbines.
Characteristic Asynchronous type Doubly fed
type
Direct-drive
type
Generator type Excitation Excitation Permanent
magnet
Fault frequency High Higher Low
System reliability Low Higher High
Influence on power grid Serious Serious Minimal
Recovery power
generation after fault
Difficult Easy Easy
Reactive absorption Yes No No
Low-voltage ride
through
Weak Strong Stronger
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
466
7. wind turbines in a wind farm.
4.5. Loss of wind power
Power plants experience a certain amount of loss in the power
generation and transmission process. Wind power loss mainly involves
the internal consumption of wind power (electric rate of comprehen-
sive field and loss resulting from equipment failure and maintenance)
and the loss resulting from limited power networks.
In the case of wind power curtailment, a wind farm is capable of
sending electricity but fails to do so because of the limitation in the
power grid transmission channel and other factors, including safety
issues. These factors do not include loss of electricity as a result of
equipment failure.
The calculation method for wind power generation in a wind farm is
released by an electrical supervisor when the output of the wind farm is
limited; the model machine method is used to calculate wind power.
4.6. Friendly wind farm conditions
To improve the positive interaction between wind farms and power
grids and the safe operation level of power networks, provincial power
grid operations and wind farm management must meet the following
requirements:
(1) Basic wind farm management
The basic management of a wind farm includes data manage-
ment, personnel management, and regulatory systems.
(2) Grid technical conditions
According to the “anti-accident measures for wind power grid
operations” and the “technology requirements for wind farm access
power systems,” wind farms must be equipped with a dynamic
reactive power compensation device and an on-line power quality
monitoring device. Moreover, wind farms must be checked for
their low-voltage ride through capability, accuracy of the relay
protection device, and equipped fault recording and high-precision
forecasting devices.
(3) Operations management
Operations management mainly involves thorough examinations
and repairs. It is best utilized when dealing with accidents and ensuring
safe operations.
According to the analysis of the factors affecting the level of wind
power utilization, an evaluation index system for wind power utiliza-
tion levels is established, as shown in Fig. 11. The proposed system
includes comprehensive factors, including the production of wind
power and the interaction between wind power and power grids, which
affect the level of wind power utilization in the process.
5. Comprehensive evaluation method
5.1. Improved analytic hierarchy process
In the improved analytic hierarchy process, a judgment matrix that
uses three scales instead of the traditional nine scales is constructed
with a high degree of convergence speed and consistency [27]. The
steps are as follows:
1. To construct a hierarchy, divide the evaluation object into a target
layer, a criterion layer, and an index layer.
2. Calculate the weight of the criterion layer as follows:
(1) Establish the comparison matrix according to the relative impor-
tance of two indicators:
A
a a a
a a a a
a a a a
a a a a
=
0 ⋯
⋯
⋯
⋮ ⋮ ⋮ ⋯ ⋮
⋯
,
n
n
n
n n n nn
1 2
1 11 12 1
2 21 22 2
1 2
⎡
⎣
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
(6)
where a1,…,an are the indexes of the criterion layer; aij can be
expressed as follows:
a
a is a
a a are e
a is a
=
2 more important than ;
1 and qually important;
0 more important than .
.ij
i
i
i
j
j
j
⎧
⎨
⎪
⎩
⎪
(7)
(2) Establish the structure judgment matrix B, i.e.
B
a a a
a b b b
a b b b
a b b b
=
0 ⋯
⋯
⋯
⋮ ⋮ ⋮ ⋯ ⋮
⋯
n
n
n
n n n nn
1 2
1 11 12 1
2 21 22 2
1 2
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
(8)
Fig. 11. Index system for wind power utilization level.
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
467
8. ∑b
c c
c c
c a=
+ 1 ≥
+ 1 <
=ij
c c
c i j
c c
c i j
i
j
n
ij
−
−
−1
=1
i j
j i
min
min
⎧
⎨
⎪⎪
⎩
⎪
⎪
⎡
⎣
⎢
⎤
⎦
⎥
c c c c= min{ , ,…, }.nmin 1 2 (9)
(3) Calculate the weight, and determine the consistency. To calculate
the maximum eigenvalue of matrix B, γmax, and the correspond-
ing feature vector C, which is the weight vector, γmax is integrated
into the consistency index I=(γmax−n)/(n−1). If I is less than 0.1,
then the judgment matrix is in conformity with the requirements.
Otherwise, the comparison matrix needs to be recalculated, and
the consistency must be checked until the requirements are met.
(4) Calculate the weight of each index. The weight vector formed from
the judgment matrix is obtained by m experts by repeating
processes ①–③, i.e.,
C c c c k= { , ,…, }, ( = 1, 2 … m)k k k
n
k( )
1 2 (10)
∑E
m
C k=
1
, ( = 1, 2 … m).
k
m
k
=1
( )
(11)
In the formula, m is the number of participants, C(k)
is the weight
vector formed by the k judgment matrix, and E is the weight vector of
expectations. Each index weight vector P can be obtained by normal-
izing weight vector E.
5.2. Comprehensive fuzzy evaluation method
5.2.1. Determining the evaluation level
According to the national standard and the feasibility report of the
wind farm, the wind power utilization level of a wind farm is divided
into five evaluation grades: D={excellent, good, moderate, qualified,
poor}, and Dk∈[0,1].
5.2.2. Building the single factor evaluation matrix
According to Fig. 11, the evaluation index system for wind power
utilization level is given as A={A1, A2, A3, A4, A5, A6}. Let the i
evaluation of a single factor be Ri=(ri1, ri2, ri3, ri4, ri5). In the
formula, rik is the degree in which the i factor belongs to the k level in
the evaluation.
R
R
R
R
R
R
R
r r r r r
r r r r r
r r r r r
r r r r r
r r r r r
r r r r r
= =
1
2
3
4
5
6
11 12 13 14 15
21 22 23 24 25
31 32 33 34 35
41 42 43 44 45
51 52 53 54 55
61 62 63 64 65
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
⎥
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
(12)
5.2.3. Determining the fuzzy evaluation set
W, the comprehensive fuzzy evaluation set for wind power utiliza-
tion levels, represents the product of the normalized weight vector P
and the single factor evaluation matrix R, i.e.,
W w w w w w w P R
p p p p p p
r r r r r
r r r r r
r r r r r
r r r r r
r r r r r
r r r r r
= ( ) = ⋅
= ( )⋅
1 2 3 4 5 6
1 2 3 4 5 6
11 12 13 14 15
21 22 23 24 25
31 32 33 34 35
41 42 43 44 45
51 52 53 54 55
61 62 63 64 65
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
(13)
In this way, the level of wind power utilization is converted to the
value of W, and wk is the degree in which the wind power utilization
level belongs to the k level in the evaluation.
5.2.4. Determining the level of the evaluation object according to the
evaluation grade
The evaluation level is a non-number; thus, it should be quantified.
The definition of the corresponding relation is shown in Table 4.
Let Q be the evaluation value of the wind power utilization level.
The weighted average method is used to determine the wind power
utilization level Q, i.e.,
Q w w w w w q q q q q= ( , , , , )⋅( )T
1
0
2
0
3
0
4
0
5
0
1 2 3 4 5 (14)
Thus, the comprehensive evaluation of wind power utilization levels
is transformed into a quantitative analysis.
6. Case application
The Xinjiang grid connected to the northwest 750 kV channel began
its operations on November 3, 2010. The Xinjiang power grid was
officially connected to the national grid during this period [28].
Hami, the output port of electric power, is the east gate of Xinjiang.
The electric power output under high pressure is shown in Fig. 12.
The A line is the AC project of Golmud–Hami at 750 kV. The B line
is ± 800 kV between Sourthern Hami and Zhengzhou. The C line is ±
1,100 kV between Northern Hami and Chongqing.
Built outside Hami–Zhengzhou, the Hami–Chongqing UHV DC
transmission project, and the Hami–Sand–Golmud 750 kV second
channel, Hami is the first to house a strong and smart power grid
and is thus a fire, wind, and light cluster power energy base.
Take for example an operating wind farm in Hami. The main
parameters according to the evaluation index system of 2014 obtained
from the Hami Statistics Bureau are shown in Table 5.
Table 4
The definition of the corresponding relation.
Grade Excellent Good Moderate Qualified Poor
Symbol q1 q2 q3 q4 q5
Quantized value 0.9 0.7 0.5 0.3 0.1
Fig. 12. UHV transmission projects in Hami.
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
468
9. (1) Determining the evaluation factor set
As shown in Fig. 11, the evaluation factor set of the first
subtarget A is A={A1, A2, A3, A4, A5, A6}.
A1={A11, A12, A13}, A2={A21, A22, A23}, A3={A31, A32},
A4={A41, A42},
A5={A51, A52, A53}, and A6={A61, A62, A63}.
The evaluation factor sets of the third subtarget A are
A61={A611, A612, A613}, A62={A621, A622, A623, A624},
and A63={A631, A632, A633}.
(2) Determining the weight set
Ten experts were invited to participate in the review of this
project. According to the feasibility report and relevant national
standards for wind farm operations, the experts provided scores on
the basis of the comparison of two indexes that represented the
importance of the various factors shown in Fig. 11. The judgment
matrix and the weight of each index are obtained using Eqs. (6)–
(11).
The judgment matrix and weight based on A.
A A1 A2 A3 A4 A5 A6 γmax Weight P I=(γmax−n)/
(n−1)
A1 1 5 3 1/
5
7 1/
3
6.40 0.14 I=0.08 & $2lt;0.1
Uniform con-
vergenceA2 1/
5
1 1/
3
1/
9
3 1/
7
0.04
A3 1/
3
3 1 1/
7
5 1/
5
0.07
A4 5 9 7 1 11 3 0.47
A5 1/
7
1/
3
1/
5
1/
11
1 1/
9
0.02
A6 3 7 5 1/
3
9 1 0.26
The judgment matrix and weight based on A1.
A1 A11 A12 A13 γmax Weight P I=(γmax−n)/(n−1)
A11 1 3 1/3 3.04 0.26 I=0.02 & $2lt;0.1
Uniform convergenceA12 1/3 1 1/5 0.1
A13 3 5 1 0.64
The judgment matrix and weight based on A2.
A2 A21 A22 A23 γmax Weight P I=(γmax−n)/(n−1)
A21 1 1/3 1/5 3.04 0.1 I=0.02 & $2lt;0.1
Uniform convergenceA22 3 1 1/3 0.26
A23 5 3 1 0.64
The judgment matrix and weight based on A3.
A3 A31 A32 γmax Weight P I=(γmax−n)/(n−1)
A31 1 3 2 0.75 I=0 & $2lt;0.1 Uniform con-
vergenceA32 1/3 1 0.25
The judgment matrix and weight based on A4.
A4 A41 A42 γmax Weight P I=(γmax−n)/(n−1)
A41 1 3 2 0.75 I=0 & $2lt;0.1 Uniform con-
vergenceA42 1/3 1 0.25
The judgment matrix and weight based on A5.
A5 A51 A52 A53 γmax Weight P I=(γmax−n)/(n−1)
A51 1 1/3 1/5 3.04 0.1 I=0.02 & $2lt;0.1
Uniform convergenceA52 3 1 1/3 0.26
A53 5 3 1 0.64
The judgment matrix and weight based on A6.
A6 A61 A62 A63 γmax Weight P I=(γmax−n)/(n−1)
A61 1 1/5 1/3 3.04 0.1 I=0.02 & $2lt;0.1
Uniform convergenceA62 5 1 3 0.64
A63 3 1/3 1 0.26
The judgment matrix and weight based on A61.
A61 A611 A612 A613 γmax Weight P I=(γmax−n)/(n−1)
A611 1 1/5 1/3 3.04 0.1 I=0.02 & $2lt;0.1
Uniform con-
vergence
A612 5 1 3 0.64
A613 3 1/3 1 0.26
The judgment matrix and weight based on A62.
A62 A621 A622 A623 A624 γmax Weight P I=(γmax−n)/
(n−1)
A621 1 1/3 1/5 1/7 4.12 0.06 I=0.04 &
$2lt;0.1
Uniform con-
vergence
A622 3 1 1/3 1/5 0.12
A623 5 3 1 1/3 0.26
A624 7 5 3 1 0.56
The judgment matrix and weight based on A63.
A63 A631 A632 A633 γmax Weight P I=(γmax−n)/(n−1)
A631 1 1/3 1/5 3.04 0.1 I=0.02 & $2lt;0.1
Uniform con-A632 3 1 1/3 0.26
Table 5
Main parameters obtained from Hami wind farm in 2014.
Type Direct-drive type Unit Cumulative Year
Wind resource Average wind speed m/s 8.21
Effective wind speed hours h 8,306.22
Average air density kg/m3
1.08
Wind power
productivity
Generating capacity MkWh 152.43
Installed capacity MW 50.5
Loss of wind power Farm consumption % 0.74
Outage loss rate % 11.1
Wind power curtailment % 5.1
Equipment
operation
Wind turbine utilization % 94.45
Repair time for cases of
equipment failure
h 3,718.12
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
469
10. vergenceA633 5 3 1 0.64
(3) Determining the evaluation set
The evaluation set is D={excellent, good, moderate, qualified,
poor}, and Dk∈[0,1].
(4) Single-factor evaluation
(a) Third-stage subtarget evaluation
The fuzzy evaluation matrix of the third-level subtargets is
obtained using the evaluation criteria based on actual wind farm
configurations, “anti-accident measures for wind power grid
operations,” and “technical requirements for wind farms connected
to power systems.”
R R R=
0.7 0.3 0 0 0
0.4 0.5 0.1 0 0
0.3 0.6 0.1 0 0
=
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
=
0.2 0.8 0 0 0
0.4 0.5 0.1 0 0
0.7 0.2 0.1 0 0
A A A61 62 63
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
⎡
⎣
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
(15)
The fuzzy evaluation matrix of the third-level subtargets is obtained
with the formula F P R= ⋅ .
R =
0.404 0.506 0.09 0 0
1 0 0 0 0
0.572 0.338 0.09 0 0
A6
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
(16)
(b) Second-stage subtarget evaluation
Combined with the given comment set for evaluate indexes A1,
A2, A3, A4, A5, and A6, the fuzzy statistical method is used to
obtain the following fuzzy evaluation matrices:
R R
R R
R R
=
0.5 0.5 0 0 0
0.3 0.7 0 0 0
0.4 0.6 0 0 0
=
0.2 0.2 0.2 0.2 0.2
0.2 0.2 0.2 0.2 0.2
0.3 0.6 0.1 0 0
= 0.7 0.2 0.1 0 0
0.4 0.3 0.1 0.2 0
=
0.2 0.4 0.2 0.1 0.1
0.1 0.5 0.3 0.1 0
=
0.6 0.3 0.1 0 0
0.6 0.2 0.1 0.1 0
0 0.2 0.4 0.4 0
=
0.4040 0.5060 0.0900 0 0
1 0 0 0 0
0.5720 0.3380 0.0900 0 0
A A
A A
A A
1 2
3 4
5 6
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
⎡
⎣
⎢
⎤
⎦
⎥
⎡
⎣
⎢
⎤
⎦
⎥
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
⎡
⎣
⎢
⎢
⎤
⎦
⎥
⎥
(17)
The fuzzy comprehensive evaluation matrix of the second-level
subtarget RA is obtained with the formula F P R= ⋅ .
R =
0.4160 0.5840 0 0 0
0.2640 0.4560 0.1360 0.0720 0.0720
0.6250 0.2250 0.1000 0.0500 0
0.1750 0.4250 0.2250 0.1000 0.0750
0.2160 0.2100 0.2920 0.2820 0
0.8297 0.1385 0.0324 0 0
A
⎡
⎣
⎢
⎢
⎢
⎢
⎢
⎢
⎤
⎦
⎥
⎥
⎥
⎥
⎥
⎥ (18)
(c) First-stage subtarget evaluation
The fuzzy evaluation vector of A is obtained with the formula
F W R= ⋅A A A, i.e.,
F = (0.4167 0.3557 0.1325 0.0590 0.0381).A
Then, the comprehensive evaluation of A is performed as follows:
Q = [0.4167 0.3557 0.1325 0.0590 0.0381]⋅(0.9 0.7 0.5 0.3 0.1)
= 0.7118
A
T
The result shows the good grade of the wind power utilization level
of the wind farm.
7. Result analysis
The results of the first-stage subtarget evaluation are shown
Table 6.
In the first level, the main factors influencing the wind power
utilization level are indexes A6, A4, and A1, which are evaluated. The
comprehensive evaluation of each index can be calculated with a fuzzy
evaluation vector:
Q Q
Q
= (0.4160 0.5840 0 0 0)
⋅(0.9 0.7 0.5 0.3 0.1) = 0.7832
= (0.1750 0.4250 0.2250 0.1000 0.0750)
⋅(0.9 0.7 0.5 0.3 0.1) = 0.6050
= (0.8297 0.1385 0.0324 0 0)
⋅ (0.9 0.7 0.5 0.3 0.1) = 0.8599
A
T T
T
1 A4
A6
Index Weight Result Grade
A1 A11 0.26 0.80 Good
A12 0.1 0.76 Good
A13 0.64 0.78 Good
A4 A41 0.75 0.60 Moderate
A42 0.25 0.62 Moderate
A6 A61 0.1 0.7628 Good
A62 0.64 0.90 Excellent
A63 0.26 0.7964 Good
On the basis of the evaluation of these indicators, the corresponding
conclusions can be drawn:
(a) The wind resources are in good condition (0.7832). In 2014, the
average wind speed of the evaluated wind farm was 8.21 m/s,
which accounted for 86.99% of its annual wind power. The wind
farm can thus be considered as a good wind resource.
(b) The power production is not high (0.6050). With the selected wind
farm being new, the unit operation necessitates a few adjustments,
especially in terms of unit fault number. The fan equipment has an
average utilization rate of only 94.45%, which is lower than the
average utilization rate of a Chinese wind turbine equipment.
Moreover, many people have abandoned wind power rationing,
which leads to low electrical energy productivity.
(c) Wind farms have certain characteristics (0.8599). Wind turbines
are large-capacity direct-drive units with full power during A-D-A
transformation and grid operation. These devices do not require
reactive power compensation, and they feature a low-voltage ride
through capability, which greatly improves the stability of wind
farm operations under certain friendly conditions. Thus, the
evaluation results are in agreement with the actual situation.
In summary, the evaluation index system and method proposed in
this study are scientific in nature and can thus serve as references in the
planning and design of wind farms.
The depth of this research is to be improved, in this paper we only
select a wind farm to study, which is not able to fully verify the wind
power farm operation comprehensive evaluation model rationality. We
should select different types of wind farm to verify, in order to improve
Table 6
the results of the first-stage subtarget evaluation.
Index Weight Result Grade
Wind resource characteristics A1 0.14 0.7832 Good
Type of wind turbine A2 0.04 0.6536 Moderate
Equipment operation A3 0.07 0.7850 Good
Wind power productivity A4 0.47 0.6050 Moderate
Wind power loss A5 0.02 0.5720 Moderate
Friendly wind farm conditions A6 0.26 0.8599 Good
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
470
11. the evaluation model. In addition, we will consider the impact of social
and environmental factors on the operation of the wind farm. In future
research, we will extend the operation of wind farm to a wider field, in
order to more comprehensive and effective evaluation of the operation
of wind farm.
Acknowledgements
This study forms part of a research project supported by the
Ministry of Education Innovation Team Project of China (No.
IRT1285), the National Natural Science Foundation of China (Nos.
51666017, 51606163), the major research projects of Xinjiang Uygur
Autonomous Region (No. 201230115-3), the Natural Science
Foundation of Xinjiang Uygur Autonomous Region (No.
2016D01C062), and the Xinjiang University Doctor Innovation
Project (No. XJUBSCX-201223), the Scientific Research Fund
Project of Xinjiang Institute of Engineering (No. 2015xgy291712).
The authors would like to express their gratitude for the support of
these funding authorities.
References
[1] Fan J, Wang Q, Sun W. The failure of China's energy development strategy 2050
and its impact on carbon emissions. Renew Sustain Energy Rev 2015;49:1160–70.
[2] Fan X, Wang W, Shi R, Li F. Analysis and countermeasures of wind power
curtailment in China. Renew Sustain Energy Rev 2015;52:1429–36.
[3] Petroleum B. BP Statistical Review of World Energy 2015. London; 2015.
[4] Fan X, Wang W, Shi R, Li F. Review of developments and insights into an index
system of wind power utilization level. Renew Sustain Energy Rev 2015;48:463–71.
[5] GWEC. Global Wind Statistics 2014, 2015.2.10, Brussels, Belgium.
[6] Li Junfeng, Cai Fengbo, Qiao Liming, Gao Hu, Wang Qixue, Tan Wenqian, et al.
China wind power review and outlook. Beijing: China Environmental Science Press;
2014.
[7] Zhao X, Ren L. Focus on the development of offshore wind power in China: has the
golden period come?. Renew Energy 2015;81:644–57.
[8] Xue B, Ma Z, Geng Y, Heck P, Ren W, Tobias M. A life cycle co-benefits assessment
of wind power in China. Renew Sustain Energy Rev 2015;41:338–46.
[9] Li X, Huang Y, Ma S, Li C, Ma X. Analysis on the Influence Factors of Wind Power
Accommodation. J Power Energy Eng 2015;3(04):162.
[10] Pei W, Chen Y, Sheng K, Deng W, Du Y, Qi Z. Temporal-spatial analysis and
improvement measures of Chinese power system for wind power curtailment
problem. Renew Sustain Energy Rev 2015;49:148–68.
[11] Kane L, Ault GW. Evaluation of wind power curtailment in active network
management schemes. IEEE Trans Power Syst 2015;30(2):672–9.
[12] He Y, Pang Y, Zhang J, Xia T, Zhang T. Feed-in tariff mechanisms for large-scale
wind power in China. Renew Sustain Energy Rev 2015;51:9–17.
[13] Bai JH, Zhang Q, Xin SX, Xie GH. Research on Key indexes Comparison of wind.
Power Util Lev Electr Power 2012;45(11):7–12.
[14] World Wind Energy Association (WWEA) . Small wind world report 2014.
Germany: WWEA; 2014.
[15] Sun S, Liu F, Xue S, Zeng M, Zeng F. Review on wind power development in China:
current situation and improvement strategies to realize future development. Renew
Sustain Energy Rev 2015;45:589–99.
[16] Guo Y, Ru P, Su J, Anadon LD. Not in my backyard, but not far away from me: local
acceptance of wind power in China. Energy 2015;82:722–33.
[17] CMA, wind and solar energy resources center. China wind and solar energy
resources bulletin 2014. 〈http://cwera.cma.gov.cn〉.
[18] Twidell J, Weir T. Renewable energy resources. Routledge; 2015.
[19] Xiao L, Wang J, Dong Y, Wu J. Combined forecasting models for wind energy
forecasting: a case study in China. Renew Sustain Energy Rev 2015;44:271–88.
[20] Zeng M, Duan J, Wang L, Zhang Y, Xue S. Orderly grid connection of renewable
energy generation in China: management mode, existing problems and solutions.
Renew Sustain Energy Rev 2015;41:14–28.
[21] He Y, Pang Y, Zhang J, Xia T, Zhang T. Feed-in tariff mechanisms for large-scale
wind power in China. Renew Sustain Energy Rev 2015;51:9–17.
[22] Yuan J, Na C, Xu Y, Zhao C. Wind turbine manufacturing in China: a review. Renew
Sustain Energy Rev 2015;51:1235–44.
[23] CWEA . Wind power brownouts preliminary statistics. Beijing: China
Environmental Science Press; 2014. p. 2015.
[24] The National Energy Administration. 〈http://www.nea.gov.cn/〉.
[25] The China Wind Energy Association. 〈http://www.cwea.org.cn/〉.
[26] Johnston L, Díaz-González F, Gomis-Bellmunt O, et al. Methodology for the
economic optimisation of energy storage systems for frequency support in wind
power plants. Appl Energy 2015;137:660–9.
[27] Sun S, Liu F, Xue S, et al. Review on wind power development in China: current
situation and improvement strategies to realize future development. Renew Sustain
Energy Rev 2015;45:589–99.
[28] China’s Energy development strategic action plan (2014–2020). The State Council,
Beijing, 20140.11.19.
R.-j. Shi et al. Renewable and Sustainable Energy Reviews 69 (2017) 461–471
471