Energy Policy xxx (xxxx) xxx

Contents lists available at ScienceDirect

Energy Policy
journal homepage: http://www.elsevier.com/locate/enpol

Policy analysis for grid parity of wind power generation in China

Xiaomin Xu a, b, *, Dongxiao Niu a, b, Bowen Xiao c, Xiaodan Guo d, Lihui Zhang a, b, Keke Wang a, b
a
School of Economics and Management, North China Electric Power University, 102206, Beijing, China
b
Beijing Key Laboratory of New Energy and Low-Carbon Development (North China Electric Power University), 102206, Beijing, China
c
School of Economics and Management, Beihang University, 100191, Beijing, China
d
School of Environment & Natural Resources, Renmin University of China, 100872, Beijing, China

A R T I C L E I N F O A B S T R A C T

Keywords: The “13th Five-Year Plan” for wind power has proposed that it will reach grid parity and compete with power
Wind power and hydropower. Accordingly, many doubts have been raised. Is the wind power in China already equipped with
Subsidy policy conditions for grid parity? What is the impact on the development of wind power? To solve these doubts, this
Wind power curtailment forecasting
study employs a system dynamics model to judge whether China can achieve grid parity for wind power. First,
Grid parity
System dynamics
the factor indicator system is constructed from the aspects of wind power production, consumption, and
curtailment. Second, the trend of wind power curtailment, cost, revenue, and installed capacity are predicted
from 2005 to 2030. Third, three scenarios are set to simulate the impact of grid parity on wind power. Empirical
results show that: (1) Net revenue and installed capacity will continue to increase, while the wind power
curtailment will gradually reduce. (2) When the subsidy is decreased to 0, revenue will significantly reduce, and
the installed capacity will reduce by nearly 1/4. (3) The Chinese government should not abolish all subsidies for
wind power to achieve grid parity in 2020. To prompt the process for the grid parity of wind power, some policy
implications are proposed.

1. Introduction Coupled with the configuration model of Internet bidding proposed by

the “518” Wind Power New Deal, the entire wind power industry is
Substantial financial subsidies have always been a potent agent for under great pressure to reduce its costs (Hache and Palle, 2019).
the government to support the rapid development of an industry. The Simultaneously, wind power development is also faced with various
wind power industry is no exception. As wind power is heavily depen­ kinds of problems, such as the failure to guarantee priority scheduling of
dent on climate, its output has characteristics of uncertainty and renewable energy, serious wind power curtailment, and the delay in the
randomness. Due to its high cost in its early stages of development, wind payment of subsidies (Fendri and Chaabene, 2019; Wang, 2018; Yan
power development initially relies heavily on government subsidies et al., 2018).
(Rosales-Asensio et al., 2019). However, subsidies are not a long-term As a widely distributed, inexhaustible, clean, and efficient renewable
solution. Recently, as the price of onshore wind turbine equipment has energy, wind energy has become the preferred source energy for low
fallen, the need to reduce wind power subsidies and set the same price carbon and sustainable development of human society (Hernandez et al.,
with thermal power has been increasingly called for (Yao et al., 2015). 2019; Sharifian et al., 2018). The development of large-scale wind
The “13th Five-Year Plan” of wind power has also proposed to reduce power projects can increase the proportion of clean energy, replace fossil
the cost of new energy generation and achieve the goal of grid parity. It energy consumption, reduce pollutant emission, and curb environ­
is expected that by 2020, the price of wind power will be able to compete mental pollution (Wang and Zou, 2018). Consequently, wind energy
with local coal power generation. In September 2017, the National En­ development has become a global trend and an important part of the
ergy Administration announced the first batch of thirteen demonstration energy strategy in many countries (Xia and Song, 2017; Zendehboudi
projects for grid parity of wind power (National Energy Administration, et al., 2018). China’s cumulative installed capacity of wind power is
2017). The thirteen projects involve Hebei, Heilongjiang, Gansu, Ning­ increasing yearly from 2011 to 2017, as shown in Fig. 1(a). Among
xia, Xinjiang, and other provinces, with a total capacity of 707,000 kW. them, the Northeast, Northwest, and North China areas (the “Three

* Corresponding author. School of Economics and Management, North China Electric Power University, 102206, Beijing, China.
E-mail addresses: xuxiaomin0701@126.com, 50602412@ncepu.edu.cn (X. Xu), niudx@126.com (D. Niu), 15353638996@163.com (B. Xiao), 15811440711@
163.com (X. Guo), zlh6699@126.com (L. Zhang), 15652912329@163.com (K. Wang).

https://doi.org/10.1016/j.enpol.2019.111225
Received 6 April 2019; Received in revised form 18 December 2019; Accepted 26 December 2019
0301-4215/© 2020 Elsevier Ltd. All rights reserved.

Please cite this article as: Xiaomin Xu, Energy Policy, https://doi.org/10.1016/j.enpol.2019.111225
X. Xu et al. Energy Policy xxx (xxxx) xxx

North” areas) are still the gathering areas for wind power installations, 2018 (Polaris Wind Power Network, January 2019). The subsidy arrears
accounting for about 74% of the total wind power installed capacity. time is about three years, which has a great impact on the operation of
After the upsurge for wind power, the pace for development has grad­ enterprises. Some enterprises fill the gap by transferring the assets of the
ually slowed. Globally, the world’s newly installed capacity of wind power station.
power decreased by 7.4% yearly in 2017. In the same year, compared Combining the background of wind power development, and based
with 2015 and 2016, the newly installed capacity of China’s wind power on existing research, this study constructs an index system of wind
declined for two consecutive years (National Renewable Energy Center; power influencing factors, uses a dynamic system dynamics model to
Renewable Energy Development Center, 2018). In 2017, the wind power analyze the promotion and restriction of various factors, and predicts
installation capacity layout in China continued to shift to the eastern and the future situation for wind power under different subsidies combining
central regions. The layout of wind power installations has been further the historical data. First, combining the influence factors of wind power
optimized, but it has increased the difficulty of engineering construc­ development, the system dynamics model is constructed from three
tion, making non-technical costs increase. Simultaneously, the approval aspects of electricity consumption, wind power generation, and wind
and environmental protection requirements are becoming increasingly power curtailment. Second, to better grasp the future consumption trend
stringent. These have caused a decline in the installed capacity of wind of wind power, this study predicts the trend of wind power curtailment.
power (Lin et al., 2016). Combined with the results, the future trend of wind power revenue and
Additionally, the wind power consumption level cannot keep up with installed capacity can be predicted. Third, due to the high dependence
the speed of wind power installation development. The rapid growth of degree on the subsidy policy, this study takes wind power subsidies as
wind power is at the expense of a severe waste of wind power resources exogenous variables, sets up different scenarios, simulates policy
in China (Shukla and Singh, 2016; Wu and Li, 2017). A phenomenon of strengths, and analyzes the effect of different scenarios on wind power
wind curtailment occurred for the first time in 2010 and peaked in 2016. development to provide decision support for the future direction of wind
Since China’s implementation of wind power investment monitoring power. Finally, some countermeasures and suggestions are proposed for
and new project management measures, China’s wind curtailment better and healthier development in the future.
phenomenon has been somewhat relieved. In 2017, the overall wind The remainder of this paper is organized as follows: Section 2 studies
curtailment was 419 billion kWh, while it decreased by 142 billion kWh literature on the impact of government subsidies on the development of
in 2018, whose wind curtailment rate declined from 12% to 7%, wind power. Section 3 introduces the applicability analysis of the model
decreased by 5 percentage points yearly (Yin et al., 2019), as shown in and describes the modeling process. Section 4 describes the empirical
Fig. 1(b). Among them, Gansu Province has the largest decline (14 analysis based on the modeling process and discusses the results. Firstly,
percentage points). However, it is still one of the areas with the worst the wind power development is predicted from the aspects of installed
wind power curtailment (Nguyen et al., 2016), seen in Fig. 2. Despite capacity, income, cost, and wind abandonment. Then the prediction
China achieving the decline in quantity and rate of wind power accuracy of the model is tested. Finally, the model is applied to simulate
curtailment in 2018, there is still a high rate in some areas and a gap the impact of different subsidies on wind power. Section 5 summarizes
between the reasonable deviations of 5%. In particular, Gansu, Xinjiang, the conclusions and provides suggestions.
Jilin, and other regions are still in the red zone of wind power invest­
ment detection and warning (State Grid Energy Research Institute Co., 2. Literature review
2018). The contradiction between new energy and power consumption
is still outstanding. Wind power curtailment has become increasingly In its early stage of development, wind power relied on various forms
prominent and the biggest obstacle to the sustainable development of of subsidies to achieve the mutual promotion of scale expansion and cost
wind power in China. reduction and enhance industrial competitiveness. However, the rapid
The arrears of subsidies affect the cash flow of wind power operators, development of wind power-driven by high subsidy policies is not sus­
which is a serious injury to wind power development. According to the tainable. Additionally, the rapid development of new energy has
feed-in tariff pricing method of new energy, the price exceeding the coal- brought enormous pressure on finances. The subsidy gap is increasing.
fired benchmark is partially covered by financial subsidies. Renewable According to statistics, the subsidy gap has already reached 100 billion
energy additions are the main source of new energy subsidies, which is yuan in 2019 and will reach 1 trillion yuan by 2030.
calculated based on the difference between the subsidy demand and the Accordingly, many domestic and foreign scholars have expressed
levy amount. However, with the sustained development of wind power, their views on whether onshore wind power can completely withdraw
the subsidy gap is growing. According to relevant statistical data, the subsidies and achieve grid parity in 2020. There are two different voices
subsidy gap has accumulated more than 100 billion yuan by the end of in the academic world. One supports the policy for wind power

Fig. 1. The cumulative installed capacity and curtailment situation of wind power in China from 2008 to 2018
Data Sources: China Renewable Energy Society Wind Energy Professional Committee (CWEA).

2
X. Xu et al. Energy Policy xxx (xxxx) xxx

Fig. 2. The wind curtailment rate of major wind power provinces in 2018
Data Sources: National Renewable Energy Center and the National Energy Administration.

integration without subsidies. They argue that substantial subsidies are Marcella and Massimo (2017) separately studied the electricity price
not a long-term solution. With the further reduction of wind power costs, subsidy policy and the tradable green certificate policy implemented by
the era of grid parity without subsidies will come sooner or later. This is the five European countries through the panel data regression model.
the inevitable law of renewable energy development worldwide (Kedron Marcella found that the electricity price subsidy policy has a positive
and Bagchi-Sen, 2018). The abolition of wind power subsidies symbol­ impact on renewable energy generation capacity and installed capacity.
izes the transition period from government support to market compe­ Zhao (2019) analyzed the path and countermeasures of grid parity from
tition. Huang (2018) analyzed 31 provinces from 2007 to 2013 and the perspective of the cost of electricity (COE) for wind power. He
concluded that the subsidy policy did not have a significant impact on believed that subsidies could not be canceled presently. Only when the
wind power development. Angus et al. (2014) found that the role of the wind power cost is close to the coal-fired desulfurization benchmark
electricity price subsidy policy is gradually decreasing, and is considered price, can the wind power realize grid parity. Yang et al. (2019) con­
to be not the most effective way to promote the development of structed a panel threshold effect model based on the sample of 92
renewable energy by analyzing the implementation effect of the UK renewable energy companies in China from 2007 to 2016. By studying
electricity price compensation policy. Zhao and Liu (2013) introduced the threshold effect of government subsidies on renewable energy in­
the SD model and showed that the quota system has a more direct and vestment, it was found that government subsidies have a positive
effective incentive effect than subsidies by taking biomass power gen­ threshold effect on renewable energy investment and are the main forces
eration as an example. Xiao and Ye (2016) analyzed the substitution and supporting the development of renewable energy enterprises.
income effect of the subsidy policy and revealed that the current subsidy Furthermore, scholars who question wind power grid parity also
policy is not the optimal policy for promoting renewable energy. Wang consider that China’s current problems of wind curtailment and subsidy
et al. (2019) believed that the current tariff subsidy policy could not arrears are still serious. These problems have become obstacles to the
meet the high-quality development requirements of renewable energy. development of wind power. Before those problems can be effectively
He used the vector autoregressive (VAR) model to study the effect of resolved, it is not suitable for wind power to reach grid parity. Currently,
electricity price subsidies on the installed capacity of renewable energy there are few studies on whether wind power generation can abolish
through impulse response function and variance decomposition. The subsidies. Most of the existing literature is limited to qualitative analysis
results showed that the cumulative effect of current electricity price that only relies on their own experience and some historical data to
subsidies is waning, and its contribution to power generation is limited. support their views. Fewer researchers use quantitative analysis to judge
The other academic opinion questions the realization of the grid policy effectiveness on wind power. The lack of quantitative calculation
parity of wind power generation in 2020. Academics with these views and model make their arguments less convincing. Quantitative analysis
believe that, despite the significant decline in the cost of wind power is involved in some literature, but it is only a simple method and a single
equipment, countries developing new energy have not eliminated the influencing factor to calculate project income (Li et al., 2016). The ac­
new energy subsidy policy (Cherp et al., 2017; Yin et al., 2018). Pres­ curacy of the results remains to be discussed. Qin et al. (2018) studied
ently, the period of wind power development has not yet had the time to the feasibility and rationality of grid parity for wind power. Through the
cancel subsidies and achieve parity (Xi, 2017). Li (2017) demonstrated analysis of the relationship among cost, utilization hours, and electricity
that technological breakthroughs, the virtuous cycle mechanism of new price, the predicted value of unit cost in 2020 is proposed according to
energy absorption, and the establishment of market mechanisms were the three scenarios of conservatism, neutrality, and optimism, which has
the key factors to realize wind power generation without subsidies, certain guiding significance for the future development of wind power
which are indispensable. China’s wind power development has not yet regionally. However, the influencing factors of wind power are
reached this condition. Guo (2018) measured the income of wind power numerous and complicated. Considering only one or two factors, the
projects through the hours wind power is used. The results showed that analysis results are still one-sided.
wind farms in most areas still cannot reach the conditions before 2020. Existing research analyzes the influence path and influence degree of

3
X. Xu et al. Energy Policy xxx (xxxx) xxx

government subsidies on the development of the wind power industry characteristics and focus on the relationship between the system struc­
from different angles, and provides reference and direction for wind ture and the dynamic behavior of the system behavior, and it can grasp
power to achieve gird parity. The literature also studies foreign renew­ the dynamic system by taking the physical system and human partici­
able energy subsidies, which has meaning for China’s wind power pation as a whole. It is believed that there is a feedback loop in the
development. However, most literature selects several indicators from a system under study. Applying the idea of system engineering, the whole
certain aspect to study the impact of subsidies on wind power. Wind system is divided into several subsystems, the causal relationships be­
power development is affected by many aspects. If we only study a tween the subsystems are clearly defined and the mathematical model of
certain aspect, we cannot reveal the internal factors and the changes in the system is established. After verifying the validity of the model, it can
wind power development. The conclusions obtained from such studies change the system conditions in a targeted way to perform multiple
are not comprehensive and convincing. System dynamics is a discipline simulation calculations to solve problems and provide the basis for
that studies the structure and behavior of complex systems. It can be formulating strategies, policies, and decisions. Accordingly, this study
used to analyze and research the information feedback system. Ac­ considers that the SD model is suitable for wind power development.
cording to the internal causal feedback characteristics of the internal The SD model can better simulate the development process of wind
components of the system, the internal structure of the system can be power systems with the advantage of dynamic simulation than other
employed to find the root cause of the problem (Wu et al., 2015). The methods can (Liu et al., 2015). Accordingly, SD is used here to simulate
model can fully consider the internal and external influence factors of and predict the future trend of wind power. To understand the change in
wind power development, analyze the effectiveness of the policy, and the path of wind power development, the promotion and restriction
obtain the wind energy price analysis results with certain reference factors affecting the development of wind power should be grasped. To
(Aslani et al., 2014). show the causality relationship between the different factors, this study
The contributions and innovations of this study are mainly reflected applies the hierarchical approach in system dynamics to find the driving
in the following four aspects from theory and application. First, a dy­ force of wind power generation (Khansari et al., 2017).
namic system model is established to analyze the development of wind Consequently, this study separates the SD model into three sub­
power in China. Second, future trends of curtailment, cost, revenue, and systems: the electricity consumption subsystem, the wind power gen­
installation capacity of wind power are predicted to grasp the direction eration subsystem, and the wind power curtailment subsystem to
of wind power development. Third, a policy scenario analysis of analyze characteristics of wind power generation development and
different subsidy intensity is conducted to investigate the effects of illuminate the relationships between the variables.
subsidy policy. Finally, policy recommendations on the development
and grid parity of wind power are suggested as a possible solution to 3.1. Framework of the model
help the Chinese government promote the development of the wind
power industry. 3.1.1. Definition of behavior boundary and time boundary
To determine the boundaries of the system under study, it is first
3. Method and model description necessary to identify the issues to be studied. The size of the problem
determines the size of the system, the variables it contains, and the
System dynamics theory (SD) is a method for system modeling and complexity of the system. Wind power is generally affected by both in­
dynamic simulation. It can analyze the dynamic complexity of social and ternal and external factors. Internal factors include the installed capacity
economic systems (Forrester, 1969, 1971). SD can help us to understand of wind power, power generation, and so on. External factors include
the time-varying behavior in the complex systems regarding system China’s economic development, population, power demand, and policy
thinking and feedback control theory (Faezipour and Ferreira, 2018). It factors.
can conduct quantitative research on complex socio-economic systems This study establishes a system for wind power development,
by the application of computer simulation technology under the prin­ focusing on the impact of various factors on it. Considering the integrity
ciple of feedback control. Since Jay W. Forrester, who had worked in the of the system, the whole of the feedback loops affecting the interaction
Massachusetts Institute of Technology (USA), founded the theory in the of wind power development and the interaction loops are taken as the
mid-1950s, its extensive use has been successful in many strategies and behavior boundary of the model, and the causal feedback relationship
decision-making analysis (He et al., 2018). The theory is known as the between internal factors is specifically explained. The system boundary
“strategy and decision-making laboratory.” A system’s internal dynamic examined in this study is determined as the change in power production
structure and feedback mechanism determine its behavior patterns and and consumption brought about by the development of wind power.
characteristics, which develops and evolves according to certain laws Therefore, the system is subdivided into three aspects: power con­
under the influence of internal and external forces and constraints. The sumption, wind power generation, and wind curtailment.
system dynamics method is a combination of qualitative and quantita­ In this study, the time boundary of the system dynamics model for
tive that is suitable for the study of data shortage problems and dealing wind power development is set from 2005 to 2030, in which 2005–2018
with complex social and economic issues (Hsu, 2012). is the time boundary of the model test, and 2019–2030 is the boundary
The system dynamics method is a very powerful mathematical tool of the model simulation prediction.
for handling complex nonlinear dynamic feedback processes (Feng et al.,
2018). Wind power development problems are suitable for simulation 3.1.2. System dynamics modeling of wind power development
modeling using system dynamics methods due to the following charac­ Assuming that wind power generation is an independent and com­
teristics: 1) Research on wind power development, such as wind power plete industry, its overall value is reflected by several aspects such as
curtailment prediction, is process-oriented. It is not limited to detail power production, power consumption, and the amount of wind
issues, and it studies the change in long-term development trends. 2) The curtailment. The power generation, that is, wind power generation,
wind power development system complies with the law of causation. represents the realization of the final benefit of wind power develop­
This kind of mutual causality is the feedback relationship in cybernetics, ment. However, it is also restrained by electricity consumption and wind
and there are multiple causal feedback loops within the limiting factors power curtailment. Therefore, considering that the production capacity
and development trends of wind power. Therefore, its structure is of wind power in China is determined not only by the power con­
complex, and its behavior expressed by the outside world is often sumption and the cost of wind power generation, but also by the amount
counter-intuitive. 3) There are complex non-linear dependencies among of wind curtailment, this study finalizes three subsystems of electricity
various elements of the wind power development system (Hassan, consumption, wind power generation and wind power curtailment to
2013). The system dynamics method can fully consider the above comprehensively measure the level and benefits of wind power

4
X. Xu et al. Energy Policy xxx (xxxx) xxx

development.
According to the causal logic relationship between different factors,
this study uses Vensim software to establish the dynamic flow diagram
of wind power development. The flow graph is applied to construct a
system dynamics overall model by constructing a causal relationship
between variables between the system variables and the control equa­
tions. This flow graph contains a total of 60 variables, including state
variables representing the cumulative results (represented by boxes),
rate variables representing the velocity of the state variables (repre­
sented by double triangles), and other auxiliary variables, as shown in
Fig. 3. To more clearly represent the logical relationship between vari­
ables, this study splits the overall system dynamics model into three
subsystems and explains the important control equations involved.

3.2. Flow diagram

3.2.1. Electricity consumption subsystem Fig. 4. The electricity consumption subsystem flow diagram.
To estimate the total electricity consumption, we consider residential
and industrial electricity consumption. Theoretically, total electricity power generation is shown in Fig. 5.
consumption can be separated into primary, secondary, and tertiary Wind power cost mainly includes construction, installation, main­
industrial electricity consumption and residential electricity consump­ tenance, and other generation costs. Among them, the construction and
tion. To predict the electricity consumption precisely, we prefer to installation costs are the most significant, which can exert great influ­
predict them from electricity intensity rather than directly from time. ence on total costs in wind power plants. Due to the harsh environmental
Hence, industrial electricity consumption is determined by the corre­ requirements of energy, the wind power industry by the close attention
sponding industrial electricity consumption intensity and industrial of all countries, wind power technology has developed tremendously,
output value. As for the residential electricity consumption, it can be and the stand-alone wind turbine capacity has increased. Moreover, the
predicted by per capita household electricity consumption and popula­ increment of large generator-transformer units’ capacity can reduce the
tion. Moreover, every industrial output value is predicted from GDP, cost of manufacture and electricity generation and improve the effi­
which is determined by the GDP growth rate. Finally, the flow diagram ciency of electricity generation. Hence, the construction and installation
of the electricity consumption subsystem is presented in Fig. 4. costs are decreased over time, whose calculation formula is shown in Eq.
(1).
3.2.2. Wind power generation subsystem
The wind power generation subsystem is the most important module construction and installation cos t ¼ wind power capacity⋅construction
in this study. In this part, there are five main aspects: cost, revenue, and installation coefficient (1)
investment, policy, and wind power capacity. The flow diagram of wind

Fig. 3. Dynamic modeling of the wind power development system.

5
X. Xu et al. Energy Policy xxx (xxxx) xxx

Fig. 5. Flow diagram of wind power generation subsystem.

In the context of the rapid development of wind power installed wind turbines under the normal circumstances, which is mainly due to
capacity, single wind turbine capacity and cost are increasing in the the lack of acceptance of the local power grid, the randomness, inter­
meantime. The maintenance cost of wind power plants still maintained mittent, and the non-schedulability of wind power output. The large
at a high level. Currently, the warranty period for wind turbines is 3–5 scale of wind power absorption remains a conundrum worldwide. The
years. After the warranty period, the maintenance cost of the wind problems in this area are more prominent in China. With the continuous
turbine passes on to power plants from manufacturers. It is estimated expansion of China’s wind power development and construction scale,
that the maintenance cost of each wind turbine is 20–30 thousand CNY/ the use of wind power has become increasingly prominent. Hence, here,
MW. The calculation formula of maintenance cost is seen in Eq. (2). the wind power curtailment is considered. The wind power curtailment
subsystem flow diagram is shown in Fig. 6.
maintenance cos t ¼ wind power capacity⋅coefficient (2)
Wind power curtailment has numerous influencing factors. Among
In this study, the price for wind power generation can be divided into them, the main factors are peak load capacity, interregional trans­
the benchmark price and new energy subsidies. The subsidy, which is mission capacity, and the new method to absorb the wind power gen­
determined by the policy factor, is a decisive variable influencing the eration, including using electricity to heat and produce hydrogen and so
revenue. The policy factor is formalized by a piecewise function of the on.
difference between the ideal proportion and actual value of wind power The peak load units contain the thermal power generating unit, gas-
generation, shown in Eq. (3). The more actual value lag behind the fired generating unit, and pumped-storage hydropower generating unit.
target, the higher the policy factor. The peak load capacity can be predicted by the three-power installed
capacity, shown in Eq. (5).
policy ¼ if then elseðgap > n1 ; p1 ; if then elseðgap > n2 ; p2 ; ⋯if then else
ðgap > na ; pa ; qÞ peak load capacity ¼ α⋅thermal power capacity
(3) þ β⋅gas fired power capacity þ γ⋅hydropower capacity (5)
Similarly, the investment coefficient is also a piecewise function Considering thermal power, for example, the thermal power
determined by net revenue; the higher the net revenue, the higher the installed capacity is determined by the increased rate of thermal power
investment coefficient. Its expression in Vensim software is set as: and the increment of thermal power. The equations are seen in Eqs. (6)
IF THEN ELSE (net revenue>1000, 0.8, IF THEN ELSE(net rev­ and (7)
enue>500, 0.6, IF THEN ELSE(net revenue>200, 0.5, IF THEN ELSE
increment of thermal powerðtÞ ¼Thermal power capacityðtÞ⋅increase rateðtÞ
(net revenue>100, 0.4, IF THEN ELSE(net revenue>0, 0.3,0))))).
The newly-added capacity has a close relationship with the invest­ (6)
ment increment, which can be obtained by dividing the investment
Thernal power capacityðtÞ ¼ Thermal power capacityðt dtÞ
increment by the cost of building installation, shown in Eq (4). The
increment of investment can be written as the sum of basic investment þ ðincrement of thermal powerÞdt (7)
and desired investment. The fixed investment means the annual constant
The wind power resources and demand in China are reverse
investment, and the desired investment denotes the investment depen­
distributed. The high voltage transmission network is cross-regional,
dent on the variable investment coefficient.
long-distance, and a large-capacity is necessary for cutting wind
newly added capacity ¼ ð1 þ investment coefficientÞ⋅fix investment power curtailment. Interregional transmission capacity is also con­
(4) strained by an increased rate of interregional transmission capacity. The
=construction and installation cos t
transmission factor can reflect the final proportion of interregional
transmission. Moreover, two main methods can stimulate the capacity of
3.2.3. Wind power curtailment subsystem
wind power use: For heating and hydrogen production, which can
The so-called wind power curtailment refers to the standstill of the

6
X. Xu et al. Energy Policy xxx (xxxx) xxx

Fig. 6. The wind power curtailment subsystem flow diagram.

directly cut the proportion of curtailment. and low-volatility. This study simulates the GDP growth rate, in which
Finally, considering the interregional transmission capacity, peak the data from 2005 to 2018 is actual data. Taken from the China Sta­
load capacity, and other purposes of abandoned wind power electricity, tistical Yearbook, the data from 2019 to 2030 is a combination of the
the total wind power curtailment can be predicted, as shown in Eq. (8). results of the predictions of the Institute of Economics of the Chinese
Academy of Social Sciences (CASS, 2012), OECD (OECD, 2012) and
wind power curtailment ¼ 1 ðpeaking load factor transmission factor
HSBC (Ward, 2011), as shown in Fig. 7. It can be seen that with the
heating hydrogenÞ⋅adjust factor increase of the economic aggregate and the slow release of the reform
(8) dividend, China’s potential GDP growth rate is characterized by a slow
decline in the coming period.
4. Results and analysis Similarly, the prediction results of other constant indicators can be
obtained. Based on them, the prediction and analysis of the remaining
4.1. Data description and parameter setting indicators are conducted through the causal relationship between
different factors within the system, combined with the formula and
Since data on the development of China’s wind power industry has conditions, as shown in the Appendix Equation List located at the end of
been measured from 2005, this study analyzes the relevant data from the paper.
2005 to 2018. The data of wind power installed capacity, gross national
product, national electricity consumption, and wind power curtailment
in these areas are introduced to support the system dynamics model of 4.2. Prediction of wind power development
wind power industry development. Table 1 describes the variables in the
model, where L represents the state variable, R represents the rate var­ 4.2.1. Wind power curtailment prediction
iable, A represents the auxiliary variable, and C is the constant. The data With the increase of the capacity of wind turbines and the expansion
for these variables are derived from the < China Statistical Yearbook>, of the scale of wind farms, the impact of wind power integration on grid
<China Energy Statistical Yearbook>, <China Electricity Yearbook>, operation has become increasingly prominent. The disorderly growth of
<Power Data Statistics Compilation> (2008–2018), and the published wind power capacity brings hidden dangers to the wind power curtail­
data from the National Energy Administration and China Renewable ment in the early stage of development.
Energy Society (CRES). Part of the data was collected from the internal “Wind power curtailment” was coined in China in 2010, which
statistics of the State Grid Corporation. spread rapidly from a sporadic phenomenon and almost became an open
In Table 1, some indicators exist as constants. They are used as fixed secret and difficulty in the wind power industry. The wind resources in
values in the calculation and need to be set in advance. The constants in China are mainly distributed in the Northeast, Northwest, and North
this study mainly include: GDP growth rate, primary industry electricity China areas, and the southeast coastal areas of Jiangsu Province,
intensity, secondary industry electricity intensity, tertiary industry Shandong Province, and so on. The main mode of wind power devel­
electricity intensity, growth rate of population, death rate of population, opment in China is large-scale centralized development, medium and
per capita residential electricity consumption, benchmark price, con­ high voltage access, and long-distance transmission. Owing to the hys­
struction and installation costs per capacity, increasing rate of interre­ teresis in the transmission lines, construction, and power network
gional power transmission capacity, increasing rate of interregional planning, the inter-area grid exchange and transmission capacity of
power transmission capacity, transmission limit factor/adjust factor/ China’s large wind power bases is relatively low. Over a long period,
peak load factor, increase rate of thermal power/gas-fired power/hy­ wind power generations in China are mainly balanced locally; the scale
dropower, and interregional power transmission proportion. The data of of the inter-regional and large-scale interconnected power grid is
2005–2018 is the actual value, and the data of 2019–2030 is the pre­ minuscule. Moreover, the wind power output fluctuates and is inter­
dicted value, which is estimated by trend extrapolation. mittent, which is determined by the randomicity and intermittency of
Using the GDP growth rate as an example, GDP is a comprehensive wind speed so that it will influence voltage quality and stability of the
indicator of economic development in a country or region. The selection power system. This restriction will restrict wind development and cause
of GDP mainly reflects the impact of economic factors on electricity “abandon wind,” which makes China’s wind power development unable
consumption, and it is considered to be one of the important reasons that to achieve its expected goals. The forecasting result of wind power
affect electricity consumption. Presently, China’s economic develop­ curtailment is shown in Fig. 8.
ment has changed from high-speed to high-quality, and the development In 2012, the proportion of wind power curtailment even reached
of GDP has also shifted from high-speed and high-volatility to low-speed 13%, far more than 9.6% and 8.6% in 2010 and 2011, respectively. The
target in “the wind power development 13th Five-Year plan” is not

7
X. Xu et al. Energy Policy xxx (xxxx) xxx

Table 1
Data collection and description.
Subsystem Variable name Type Unit

Electricity consumption subsystem � GDP L Billion Yuan

� GDP growth rate C %
� GDP increment R Billion Yuan
� Output value of primary industry A Billion Yuan
� Output value of secondary industry A Billion Yuan
� Output value of tertiary industry A Billion Yuan
� Primary industry electricity intensity C kWh/yuan
� Secondary industry electricity intensity C kWh/yuan
� Tertiary industry electricity intensity C kWh/yuan
� Primary industry electricity consumption A Billion kWh
� Secondary industry electricity consumption A Billion kWh
� Tertiary industry electricity consumption A Billion kWh
� Population L Ten thousand persons
� Population growth R Ten thousand persons
� Population death R Ten thousand persons
� Growth rate of population C %
� Death rate of population C %
� Per capita residential electricity consumption C kWh/person
� Residential electricity consumption A Billion kWh
� Total electricity consumption A Billion kWh
Wind power generation subsystem � Wind power capacity L Ten thousand kilowatts
� Wind power generation A Billion kWh
� Newly-added capacity R Ten thousand kilowatts
� Design value of wind power generation A Billion kWh
� Benchmark price C yuan/kWh
� Revenue/Net revenue A Billion Yuan
� Investment L Billion Yuan
� Increment of investment R Billion Yuan
� Fixed investment A Billion Yuan
� Investment coefficient A –
� Subsidy A Billion Yuan
� Cost/maintain cost/other cost/construction and installation cost A Billion Yuan
� Construction and installation costs per capacity C Ten thousand yuan/kilowatts
� Actual proportion of wind power A %
Wind power curtailment subsystem � Interregional transmission capacity L Ten thousand kilowatts
� Newly-added transmission capacity R Ten thousand kilowatts
� Increasing rate of interregional power transmission capacity C %
� Interregional power transmission demand A Billion kWh
� Transmission limit factor/adjust factor/peak load factor C –
� Wind power curtailment A Billion kWh
� Peak load capacity A Ten thousand kilowatts
� Thermal power installed capacity L Ten thousand kilowatts
� Gas-fired power installed capacity L Ten thousand kilowatts
� Hydropower installed capacity L Ten thousand kilowatts
� Increment of thermal power/gas-fired power/hydropower R Ten thousand kilowatts
� Increase rate of thermal power/gas-fired power/hydropower C %
� Interregional power transmission proportion C %

installed capacity and grid-connected capacity, but policy adjustment. 4.2.2. Cost & revenue prediction
Alternatively, more explicitly, it aims to solve the situation of “wind The wind power cost mainly includes the construction, maintenance,
power curtailment,” accompanied with enough financial support, to installation, and other generation costs. Among them, the variable costs
ensure a reasonable profit and coordinate the dispatching operation of mainly include maintenance and other generation costs. The particu­
electric power systems. Hence, the situation of “wind power curtail­ larity of wind power lies in that there are no fuel costs, so the mainte­
ment” will obtain immediate relief, which will decrease from nearly nance cost is the most significant variable cost in wind power
15% to 6% at an average rate of 1.8 percent during the 13th Five-Year. generation. It is noteworthy that the warranty period of wind turbines is
From then, although the wind power curtailment may fluctuate, it will 3–5 years, which means that the manufacturers shall bear the mainte­
remain low for several years. Finally, the average wind power curtail­ nance costs in the first several years. Once the warranty period expires,
ment maintains at around 5% during 2005–2030. the manufacturers will pass all the maintenance costs on to power plant.
Fig. 9 depicts that there is an exponential increase in the mainte­
nance cost, which reflects a dilemma in wind power development. Given

8
X. Xu et al. Energy Policy xxx (xxxx) xxx

power industry began to accelerate development. The Chinese govern­

ment has proposed the non-fossil energy goal that 15% of non-fossil
energy in China (8% of which would have been accomplished by hy­
dropower and 3% of which would have been accomplished by wind
power). Recently, wind power outweighed nuclear power and became
China’s third-largest source of electricity, excluding hydropower and
thermal power. Although wind power development has limitations,
especially in wind power curtailment, its momentum is still strong. The
installed capacity and increased rate of wind power are shown in
Table 2.
Fig. 11 depicts the past, current, and future conditions of wind power
installed capacity graphically. The wind power installed capacity will
reach 317 GW at an average increase rate of 24%. The development
pattern can be divided into three obvious stages: (1) in 2005–2010,
driven by the promotion policy, the development of wind power gen­
eration experienced a breakneck growth with a 55% increase rate. (2) in
2010–2020, constrained by “wind power curtailment,” development
will slow down. The average increase rate will plunge from over 50% to
Fig. 7. Estimation results of China’s GDP growth rate in 2005–2030. nearly 15%. (3) in 2020–2030, although the status of “wind power
curtailment” will be alleviated, the large installed capacity will lead to a
relatively low increase rate. The average annual growth rate will
maintain at 17%. Briefly stated, there is a broad space for wind power in
the near future.

4.2.4. Validity test

To verify the validity of the model, it is necessary to ensure that it can
accurately reflect the actual situation when conducting system analysis,
prediction, and decision-making. Therefore, this study needs to test the
effectiveness of the system dynamics simulation of wind power devel­
opment in China.
The reliability test is applied to prove the effectiveness of the system
dynamics model structure based on the reliability of the variable error
model between true values and the simulated values of multiple vari­
ables from 2005 to 2018. Four indicators, including wind power
installed capacity, revenue, cost, and wind power curtailment, are
selected as test variables. If the absolute value of the error rate of the
Fig. 8. The forecasting result of wind power curtailment. variable can be controlled within 10%, the model has a certain reli­
ability and passes the effective test.
the current status of wind power plants in China, the average mainte­ The reliability test results of the relevant indicators are shown in
nance cost is 50 thousand per MW annually. The roaring maintenance Table 3 and Fig. 12.
cost creates great pressure for wind power plants and severely restricts For RSME, MAPE, MAE, and relative error, the lower the index value,
the trend and space of wind power industry development. the higher the prediction accuracy. It can be seen that the relative error
From 2005, China’s wind power began to experience fast growth. of the test variables is controlled within 10%, which meets the re­
Notably, with the blowout-style development of the industry, the quirements of the reliability test. Therefore, the system dynamics model
amount of wind power generators surpassed 90 thousand. The “China constructed for wind power development in this study is effective and
wind power operation and maintenance market report” indicated that, can reflect the industrialization of wind power technology in China, and
in 2016–2017, the warranty period of 14GW–16GW wind power gen­ it can be used to predict the development of wind power in China under
erators will expire. By the end of 2022, it will reach 187 GW. different subsidy policies.
In 2005, wind power is still at its preliminary stage; the installed
capacity remains steady and relatively low. Although subsidies for 4.3. Subsidy policy scenario simulation
construction and electricity generation in the wind power industry were
plentiful, the high construction cost and immature technology resulted 4.3.1. Scenarios setting
in low profit in most power plants. The net revenue of wind power Due to the certain subjectivity in the parameters setting process of
generation is shown in Fig. 10. After 2010, the “wind power curtail­ transfer functions, the system simulation results are changed under
ment” has further worsened their already difficult situations. The net different key parameter settings. The subsidy is selected as a key variable
revenue turned negative in the year 2010, and, in subsequent years, the after considering the vital influencing factors for the wind power
situation has not improved. By the end of 2017, the net revenue will be installed capacity. Single-factor sensitivity analysis analyzes the influ­
positive and increase yearly. This due to 1) “wind power curtailment” ence of the different proportion of subsidies on wind power installed
peaking in 2015 and gradually decreasing with some fluctuations; 2) capacity development.
with the decreasing construction cost and mature technology, the high In 2015, the NDRC (National Development and Reform Commission)
cost is no longer a severe issue in wind power development; 3) with the issued the “notice on improving the electricity benchmark price of the wind
steady increase in installed capacity, growing electricity generation will power and solar power generation,” which cleared the trend of the
lead to considerable profits. benchmark price cut. According to the target proposed by the NEA
(National Energy Administration), the connection to the grid at an equal
4.2.3. Installed capacity prediction price between wind power generation and traditional thermal power
After China implemented the “Renewable Energy Law,” the wind generation should be achieved by the end of 2020.

9
X. Xu et al. Energy Policy xxx (xxxx) xxx

Fig. 9. The maintenance cost of wind power.

Fig. 11. The installed wind power capacity and growth rate.

Under the SS scenario, we set the subsidy reduction rate at 50%, that is,
the government will withdraw 50% of the subsidy in the year 2020.
Scenario 3 (Unsubsidized scenario, US scenario): To actively
respond to the grid parity of wind power generation, we have set the rate
of subsidy to the slope to 100%, that is, the government will withdraw
the entire subsidy in the year 2020.

4.3.2. Simulation result and analysis

According to the wind power development model constructed in
Section 3.2, the key factors affecting wind power development are
predicted, and the variables in different scenarios are combined to
explore whether wind power can achieve gird parity in 2020 online
under different scenarios. Combined with the three scenario settings, the
net revenue and installed capacity results predicted by China’s wind
power development in 2010–2030 are shown in Table 4 and Figs. 12 and
Fig. 10. The net revenue of wind power.
13.
Table 4 reflects the sensitivity between the net revenue, installed
Table 2 capacity, and subsidy. Note that scenario 3 can perfectly match the
The simulation results of installed capacity. target proposed by the NEA which will withdraw all subsidies in 2020.
Year Installed Increase Year Installed Increase Moreover, the 50% cut in subsidies is also considered.
capacity (GW) rate capacity (GW) rate Figs. 13 and 14 reflect the sensitivity between the net revenue,
2005 1.26 137.42% 2018 46.50 16.85% installed capacity, and subsidies. Scenario 3 can perfectly match the
2006 3.00 66.80% 2019 54.34 18.67% target proposed by the NEA, which will withdraw all subsidies in 2020.
2007 5.01 46.55% 2020 64.48 17.94% Moreover, the 50% cut in subsidies is also considered.
2008 7.34 36.83% 2021 76.05 18.46%
Fig. 13 reflects the sensitivity between subsidies and net revenue. It is
2009 10.04 31.09% 2022 90.09 17.85%
2010 13.16 20.98% 2023 106.17 17.32% noteworthy that the subsidies and revenue are highly correlated.
2011 15.92 19.16% 2024 124.56 16.85% Consider the net revenue in 2030 as an example, in scenario 2, the net
2012 18.97 17.74% 2025 145.55 16.43% revenue decreased by 30% compared with the base scenario. Addi­
2013 22.33 16.58% 2026 169.46 16.05%
tionally, in scenario 3, the net revenue decreased by 75% compared with
2014 26.04 20.32% 2027 196.66 17.68%
2015 31.33 14.70% 2028 231.43 17.31%
the base scenario. From the perspective of net revenue, 2020 is not the
2016 35.94 14.04% 2029 271.49 16.97% right time to withdraw the entire subsidy. The power plant will suffer
2017 40.98 13.47% 2030 317.56 16.56% great profit loss especially when the government stops the subsidy.
Fig. 14 shows the results of variations of installed capacity under
different scenarios. It is conducive for us to tell the impact of variation of
To measure the impact of the lowering of subsidies in 2020, and
subsidies on the installed capacity. It can be seen that, by the end of
more explicitly make the explanations between net revenue, installed
2030, the installed capacity in scenarios 1–3 will reach 317 GW, 267
capacity, and subsidies, a scenario analysis is applied here. The three
GW, and 236 GW separately. Compared with the base scenario, the
scenarios are:
installed capacity in scenario 2 will decrease by 15.7% and decrease by
Scenario 1 (BAU scenario): Under the BAU scenario, the develop­
25.4% in scenario 3. Briefly stated, although the sensitivity of installed
ment pattern is stable, and the subsidy will maintain as usual. Combined
capacity is weaker than the net revenue, withdrawing the subsidy will
with the changing law of historical data, the current state’s wind power
also sacrifice nearly a quarter of installed capacity. Thus, unless there is
development plan is used as a benchmark for comparison with other
a major technological breakthrough, it is not appropriate for the Chinese
scenarios.
government to withdraw all the subsidies for wind power by 2020.
Scenario 2 (Semi-subsidized scenario, SS scenario): The Chinese
Both net revenue and installed capacity have a high correlation with
government has successively introduced numerous policies to promote
subsidies. With the reduction of subsidies, net revenue and installed
the healthy development of wind power. Following the National Energy
capacity have shown different downward trends. However, the sensi­
Administration’s Notice on “Actively Promoting Non-subsidized
tivity of the two variables is reflected in different dimensions. Net rev­
Internet Access for Wind Power and Photovoltaic Power Generation,”
enue appears more sensitive to subsidies than installed capacity in
the call for grid parity of China’s wind power in 2020 is increasing.
horizontal correlation. In scenario 2, when subsidies are reduced by

10
X. Xu et al. Energy Policy xxx (xxxx) xxx

Table 3 government will not be able to withdraw all subsidies for wind
Error index of forecasting results. power by 2020.
Error index Installed Revenue Cost Wind power
capacity curtailment 5.2. Policy implications
RSME 0.021 0.019 0.012 0.03
MAPE 0.08 0.049 0.022 0.087 Currently, although China’s wind power development has achieved
MAE 0.012 0.008 0.003 0.02 remarkable results, there are also prominent problems. Further large-
Relative 6.25 4.33 5.38 9.08
scale promotion of grid parity is confronted with many tricky prob­
error
lems. From the scenario simulation and analysis, the development of
wind power depends more on subsidies. If the subsidies are canceled, it
50%, net revenue shows a 36.83% reduction, whose impact degree is will bring a greater impact on the revenue of wind power. China’s
73.66%. While, in scenario 3, when subsidies are reduced to 0, net onshore wind power is slow to usher in the era of parity grid, mainly due
revenue decreased by up to 96.34%, whose impact degree is close to to the following resistance. First, concerning wind power generation
100%. Similarly, the largest impact degree of installed capacity on efficiency, it is necessary to solve the problem of wind curtailment in the
scenario 2 and 3 is 31.32% and 25.41%. In other words, the current whole country and ensure that wind power is fully consumed to the grid.
revenue of China’s wind power industry mainly depends on subsidies. Second, from the regional perspective of wind power development, the
Once all subsidies have been removed, the wind power industry will be wind power industry began to look for opportunities in the low-speed
severely affected. regions of the central and eastern regions. The early-stage develop­
Installed capacity is more sensitive than net revenue in the longitu­ ment cost in these areas is 20% higher than in the “three north” areas,
dinal correlation. As time changes, the sensitivity of installed capacity whose grid parity for wind power is almost incompetent. Simulta­
gradually increases in the same situation. In scenario 2, the rate of neously, the lack of market mechanisms and safeguard policies to pro­
reduction in installed capacity increased from 0 to 15.66%, and the mote grid parity cannot enhance the competitiveness of wind power
degree of impact increased from 0 to 31.32%. While the net revenue after the elimination of subsidies.
shows irregular volatility, where sensitivity to subsidies in 2030 is less The imbalance of wind power resources in China leads to different
than in 2022. In other words, we cannot arbitrarily determine the time problems in different regions. Hence, the development of wind power
of the grid parity for wind power. We should comprehensively examine cannot adopt the “one size fits all” model, and the exit of wind power
its time relevance to the installed capacity. subsidies should also be gradual due to different types of wind power.
Offshore wind power faces the most technical difficulties. The current
5. Conclusions and policy implications standards, systems, and management are not in place. The construction
cost remains high, at about twice as much as onshore wind power. Grid
5.1. Conclusions parity is not possible in a short period, and certain subsidies are needed
to support technological progress and scale development. Onshore wind
To deal with many issues brought by resource crisis and environ­ power is divided into centralized wind power and distributed wind
mental pollution, new energy has become the focus of attention in many power. For centralized wind power, a large scale of development has
countries, especially in wind energy resources. Although wind power in been formed. It is necessary to make technological innovation and
China plays a significant role, wind power development bristled with market-oriented mechanism innovation in leading in achieving grid
difficulties. Accordingly, this study simulates China’s wind power parity in areas with superior resources. Distributed wind power has the
development during 2005–2030 and establishes a scenario analysis for characteristics of a small installed capacity, high unit cost, and poor
subsidy policy using a system dynamics model to explore the policy economic efficiency. Its development speed is lagging, which requires
effect on the wind power industry. The simulation results indicate that: the country to increase policy support further, expand decentralized
wind power pilots, strengthen safeguard policies, and promote the
(1) Due to the various limitations, mainly wind power curtailment, development of decentralized wind power and local consumption.
andthe relatively high cost, wind power development is The development of wind power also needs to consider the charac­
restricted. Wind power mainly depends on wind energy, which is teristics of different regions and gradually achieve grid parity in tasks.
random and unstable. When the wind power reaches a certain Areas with abundant wind resources and a strong capacity to absorb
condition, there is power output; otherwise, there is no power should consider these characteristics to conduct pilot research on grid
output, which leads to the discontinuity and instability of wind parity. Studied need to focus on how to overcome technical difficulties
power output. It is one of the leading causes of the phenomenon and reduce costs. For areas with severe wind curtailment, it is important
of wind power curtailment. Additionally, the high cost of land to address the problem of abandoned wind and expand consumption
acquisition and maintenance, especially in remote mountainous methods. For areas with insufficient wind resources and a strong ca­
areas, makes the benefits of wind power projects not considerable pacity for consumption, it is necessary to reduce non-technical costs.
and brings some resistance to the development of wind power. Additionally, the state needs to improve the green card and quota system
(2) The wind power curtailment is a severe problem on the devel­ policies and accelerate the construction of the electricity spot market.
opment path. Great efforts will be made to ease this epidemic in To solve the current problems faced by wind power, and guide
the “13th five-year.” Only through continuous construction and different types of wind power to escape subsidies and achieve grid parity
improvement of the electricity market and the use of market gradually, it is necessary to change the existing policy mode, and
mechanisms can we achieve a virtuous cycle mechanism and transfer the focus from subsidies to the market. The specific policy
reduce power curtailment for new energy power generation from recommendations are:
project development to operation and consumption.
(3) The subsidy is a primary factor in reducing the cost of wind power (1) Technological innovation and lower initial investment costs are
plants and attract more investment. Without the subsidy, most key to achieving wind power grid parity. The experience of
wind power plants will suffer great profit loss. From the scenario foreign industry development tells us that the initial investment
results, the net revenue decreased by 30% in scenario 2 and in wind power accounts for a considerable proportion of the total
decreased by 75% in scenario 3 compared with the base scenario. investment, in which the cost of wind turbines takes the vast
Unless there is a major technological breakthrough, the Chinese majority. When the cost of wind turbines is less than 70% of the
initial investment, the cost of wind power can be significantly

11
X. Xu et al. Energy Policy xxx (xxxx) xxx

Fig. 12. Error index calculation results.

Table 4
The results of the sensitivity analysis.
Year Net revenue (108 CNY) Installed capacity (GW)

0% 50% Rate 100% Rate 0% 50% Rate 100% Rate

2020 283.02 283.02 0.00% 283.02 0.00% 64.48 64.48 0.00% 64.48 0.00%
2021 665.48 475.19 28.59% 158.05 76.25% 76.05 76.05 0.00% 76.05 0.00%
2022 541.31 341.94 36.83% 19.82 96.34% 90.09 88.34 1.95% 87.46 2.92%
2023 830.48 579.28 30.25% 181.80 78.11% 106.17 101.18 4.70% 97.25 8.40%
2024 865.16 604.62 30.11% 184.99 78.62% 124.56 116.60 6.39% 111.08 10.82%
2025 922.52 753.27 18.35% 250.27 72.87% 145.55 133.91 8.00% 126.48 13.10%
2026 926.51 668.34 27.86% 268.40 71.03% 169.46 154.70 8.71% 143.59 15.26%
2027 1129.32 826.80 26.79% 319.23 71.73% 196.66 178.16 9.41% 162.60 17.32%
2028 1192.62 822.31 31.05% 240.00 79.88% 231.43 204.60 11.59% 185.29 19.94%
2029 1617.03 1148.46 28.98% 452.11 72.04% 271.49 234.36 13.67% 208.83 23.08%
2030 1825.79 1279.75 29.91% 449.48 75.38% 317.56 267.82 15.66% 236.88 25.41%

Fig. 13. The sensitivity of net revenue.

Fig. 14. The sensitivity of installed capacity.
controlled. While in China, the initial investment proportion oc­
cupies about 85%, with a large space for reduction. Despite this, enthusiasm, and tax reduction and policy subsidies for indepen­
China has certain independent research and development expe­ dent R&D enterprises. The government also ought to encourage
rience and capabilities, and high-end fans and components rely local and social capital to enter the field of wind power equip­
on imports. The resulting tariffs and technology transfer costs are ment manufacturing to mobilize the enthusiasm of independent
difficult to avoid, which directly leads to the difficulty of cost research and development, overcome the problems of localiza­
reduction. To address this issue, from the national level, the in­ tion of high-end wind turbines, maximize cost control, reduce the
dustrial layout should be uniformly deployed and integrated. proportion of wind turbines in wind power systems and reduce
Concentrating on advanced state negotiations with wind turbine the overall cost of wind power systems.
manufacturing will likely reduce the difficulty of technology (2) Non-technical costs have become an important factor influencing
introduction, strive to reach out to foreign core technologies, the pace of grid parity. The non-technical cost, including land
increase the introduction of core technologies, and reduce the book fees, rents, and transaction costs, etc., accounts for
cost of technology introduction. Regarding government policy, approximately 5%–15% of the initial investment, which em­
enterprises are encouraged to launch R&D and manufacturing bodies the interests of all parties. In the process of project

12
X. Xu et al. Energy Policy xxx (xxxx) xxx

identification and development, we should prioritize the use of Author contributions

state-owned unused land, optimize the floor space, make good
use of relevant laws and policies to reduce unreasonable devel­ X.X.M., N.D.X., X.B.W. and G.X.D. designed the experiments; X.X.M.
opment costs, reduce the cost of loans through green financial organized the paper framework, analyzed the data, wrote and revised
services, and promote the realization of grid parity. the manuscript; N.D.X. provided guidance on manuscript preparation
(3) The phenomenon of wind abandonment is the primary stubborn and supervised the whole work; X.B.W. and G.X.D. conducted the ex­
disease on the road of grid parity promotion. China shows an periments and wrote the manuscript; Z.L.H. discussed and modified the
uneven distribution between areas with superior wind resources manuscript; W.K.K. checked the language.
and areas with consumption capacity. The key to the solution is to
increase consumption access. On the one hand, locally, the gov­ Declaration of competing interest
ernment should increase policy guidance, fully develop resource
advantages, and increase the demand for electricity in the region The authors declare that they have no known competing financial
by introducing capital, building factories, or developing in­ interests or personal relationships that could have appeared to influence
dustries. On the other hand, the construction of transmission the work reported in this paper.
pipelines for wind power should be strengthened to realize long-
distance power transmission, optimize resource allocation, and Acknowledgments
achieve regional economic advantages. Furthermore, the micro-
grid should be formed in combination with regional advan­ The authors would like to acknowledge the financial support from
tages. Combining various energy modes of solar energy, tidal the National Natural Science Foundation of China (Grant No.71804045)
energy, and geothermal energy, the micro-grid network is formed and the 2018 Key Projects of Philosophy and Social Sciences Research,
to realize the “power generation-storage-electricity” circulating Ministry of Education, China (Grant No.18JZD032). This paper is also
power consumption mode for the maxim use of wind power. supported by the 111 Project (B18021), and the Fundamental Research
(4) The implementation of the “multi-agent mandatory quota” and Funds for the Central Universities (2018ZD14).
“renewable energy green certificate trading system” is the main
driving force for the grid parity of wind power. Market effec­ Appendix. Equation list
tiveness can be increased through quota constraints. According to
the power transmission capacity of the power grid and the ca­ Electricity Consumption Subsystem
pacity of the system, combined with various benchmark prices,
and reference to the principle of permitted cost-plus reasonable (1) INITIAL TIME ¼ 2005, FINAL TIME ¼ 2030, Units: Year
income, the number of guaranteed acquisition hours of wind (2) GDP ¼ INTEG (increment of gdp,184937)
power projects is verified and enforced. To ensure the imple­ (3) GDP growth rate ¼ WITH LOOKUP (Time,([(2005,0)-
mentation effect, the renewable energy quota system must be (2030,0.2)],(2005,0.1135),(2006,0.1269),(2007,0.142),
issued to local governments and grid companies, power distri­ (2008,0.0962),(2009,0.0924),(2010,0.1063),(2011,0.0949),
bution companies, and large users directly involved in the mar­ (2012,0.0775),(2013,0.0769),(2014,0.074),(2015,0.072),
ket. Through the combination of the quota system and the green (2016,0.07),(2017,0.067),(2018,0.066),(2019,0.064),
certificate trading system, the government and the market can be (2020,0.06),(2021,0.057),(2022,0.055),(2023,0.053),
effectively coordinated, the cost of wind power generation can be (2024,0.051),(2025,0.05),(2026,0.047),(2027,0.045),
reduced, the speed of wind power subsidies can be accelerated, (2028,0.042),(2029,0.04),(2030,0.038)))
and the process of grid parity will be promoted. (4) increment of gdp ¼ GDP*GDP growth rate
(5) Accelerating the construction of the power spot market is a (5) population ¼ INTEG (population growth-population death,
powerful guarantee for realizing the grid parity of wind power. 130756)
Wind power, with a lower marginal cost, has the competitive (6) population death ¼ population*population death rate
advantage of spot trading. Through the construction of the spot (7) population death rate ¼ WITH LOOKUP (Time,([(0,0)-
market, it is beneficial to find the price of electricity, optimize the (4000,10)],(2005,0.0065),(2006,0.0068),(2007,0.0069),
allocation of resources, and finally promote wind power grid (2008,0.007),(2009,0.007),(2010,0.0071),(2011,0.0071),
parity. Through the construction of a new energy spot trading (2012,0.0071),(2013,0.0072),(2014,0.0072),(2015,0.0074),
market across regions, pilot research will be conducted to pro­ (2016,0.0072),(2017,0.0076),(2018,0.0076),(2019,0.0076),
mote the export of surplus, new energy power generation, make (2020,0.0078),(2021,0.0076),(2022,0.0074),(2023,0.0072),
full use of the available transmission functions of the trans- (2024,0.007),(2025,0.007),(2026,0.007),(2027,0.0068),
regional channel, and organize the daily and intra-regional de­ (2028,0.0068),(2029,0.0067),(2030,0.0065)))
livery transactions in a market-oriented manner. The transaction (8) population growth ¼ population*population growth rate
mode is adopted to consider the channel security constraint, and (9) population growth rate ¼ WITH LOOKUP (Time,([(0,0)-
the transaction is completed through centralized bidding and (4000,10)],(2005,0.0124),(2006,0.01209),(2007,0.0121),
hierarchical clearing. The spot market is compatible with the (2008,0.01214),(2009,0.01195),(2010,0.0119),(2011,0.01193),
characteristics of new energy volatility and randomness, which is (2012,0.0121),(2013,0.0121),(2014,0.012),(2015,0.012),
conducive to expanding the space for new energy consumption. (2016,0.0118),(2017,0.0116),(2018,0.0114),(2019,0.0112),
(2020,0.011),(2021,0.01),(2022,0.01),(2023,0.009),
In summary, China’s current wind power is not suitable for the full (2024,0.009),(2025,0.009),(2026,0.008),(2027,0.008),
implementation of grid parity. The follow-up of the wind power industry (2028,0.008),(2029,0.007),(2030,0.007)))
still needs to be driven by the positive direction of policies, expanding (10) primary industrial electricity consumption ¼ primary industrial
wind power consumption, standardizing curtailment management and output value*first industry electricity intensity
investment policies to improve the power generation efficiency of wind (11) primary industry electricity intensity ¼ WITHLOOKUP(Time,
power and reduce development costs and achieve the goal of full-scale ([(2005,0)-(2030,0.02)],(2005, 0.025), (2006, 0.025),
grid parity for wind power. (2007,0.025),(2008,0.025),(2009,0.025),(2010,0.025),
(2011,0.025),(2012,0.02),(2013,0.02),(2014,0.02),(2015,0.01),
(2016,0.01),(2017,0.01),(2018,0.01),(2019,0.01),(2020,0.005),

13
X. Xu et al. Energy Policy xxx (xxxx) xxx

(2021,0.005),(2022,0.005),(2023,0.005),(2024,0.005), (39) increment of investment ¼(1 þ investment coefficient) *fix

(2025,0.005),(2026,0.005),(2027,0.005),(2028,0.005), investment
(2029,0.005),(2030,0.005)) (40) investment ¼ INTEG (increment of investment, 1000)
(12) primary industrial output value ¼ GDP*0.05 (41) investment coefficient ¼ IF THEN ELSE(net revenue>1000, 0.8,
(13) second industrial electricity consumption ¼ second industrial IF THEN ELSE(net revenue>500, 0.6, IF THEN ELSE(net rev­
output value*second industry electricity intensity enue>200, 0.5, IF THEN ELSE(net revenue>100, 0.4, IF THEN
(14) second industrial output value ¼ GDP*0.5 ELSE(net revenue>0, 0.3,0)))))
(15) second industry electricity intensity ¼ WITH LOOKUP (Time,
([(2010,0)-(2030,0.2)],(2005,0.2),(2006,0.2),(2007,0.18),
Wind power curtailment subsystem
(2008,0.18),(2009,0.17),(2010,0.17),(2011,0.16),(2012,0.15),
(2013,0.15),(2014,0.14),(2015,0.14),(2016,0.13),(2017,0.13),
(42) For heating ¼ A FUNCTION OF(Time)
(2018,0.13),(2019,0.12),(2020,0.12),(2021,0.12),(2022,0.11),
(43) For hydrogen production ¼ A FUNCTION OF(Time)
(2023,0.11),(2024,0.11),(2025,0.11),(2026,0.1),(2027,0.1),
(44) adjust factor ¼ WITH LOOKUP (Time,([(2010,0)-(2030,0.5)],
(2028,0.1),(2029,0.1),(2030,0.1),(2031,0.09),(2032,0.09)))
(2005,0),(2006,0),(2007,0),(2008,0),(2009,0),(2010,0.21),
(16) tertiary industry electricity consumption ¼ tertiary industrial
(2011,0.19),(2012,0.29),(2013,0.22),(2014,0.18),(2015,0.35),
electricity intensity* tertiary industrial output value
(2016,0.28),(2017,0.3),(2018,0.25),(2019,0.06),(2020,0.18),
(17) tertiary industry electricity intensity ¼ WITH LOOKUP (Time,
(2021,0.12),(2022,0.25),(2023,0.16),(2024,0.19),(2025,0.184),
([(2005,0)-(2030,0.025)],(2005,0.025),(2006,0.025),
(2026,0.2),(2027,0.19),(2028,0.306),(2029,0.21),
(2007,0.024),(2008,0.024),(2009,0.023),(2010,0.023),
(2030,0.284)))
(2011,0.021),(2012,0.02),(2013,0.02),(2014,0.02),(2015,0.02),
(45) Gas-fired power installed capacity ¼ INTEG (increment of gas-
(2016,0.02),(2017,0.02),(2018,0.02),(2019,0.02),(2020,0.02),
fired power,19)
(2021,0.02),(2022,0.02),(2023,0.02),(2024,0.02),(2025,0.02),
(46) hydropower installed capacity ¼ INTEG (increment of
(2026,0.019),(2027,0.019),(2028,0.019),(2029,0.019),
hydropower,102)
(2030,0.019),(2031,0.018),(2032,0.018)))
(47) Thermal power installed capacity ¼ INTEG (increment of thermal
(18) tertiary industrial output value ¼ GDP*0.45
power,370)
(19) per capita electricity consumption ¼ 32.542*(Time-2004)þ
(48) increase rate of gas-fired ¼ 10^(-5)*(Time-2004)^3–0.0005*
195.43
(Time-2004)^2 þ 0.0053*(Time-2004)þ0.0877
(20) total electricity consumption¼((primary industrial electricity
(49) increase rate of hydropower ¼ 0.3669*(Time-2004)^(-0.95)
consumption þ second industrial electricity consumption þ third
(50) increase rate of thermal power ¼ 0.317*(Time-2004)^(-0.946)
electricity) þ residential electricity consumption)*10000/100
(51) increment of gas-fired power ¼ Gas-fired power installed capac­
ity*increase rate of gas-fired
Wind Power Generation Subsystem
(52) increment of hydropower ¼ hydropower installed capacity*in­
crease rate of hydropower
(21) wind power generation ¼ design value of power generation*(1-
(53) increment of thermal power ¼ increase rate of thermal power*­
wind power curtailment)
Thermal power installed capacity
(22) design value of power generation ¼ design value of hours*Wind
(54) increase rate of interregional transmission capacity ¼ WITH
power generation capacity
LOOKUP(Time,([(2010,0)-(2030,0.2)],(2005,0.2),(2006,0.2),
(23) design value of hours ¼ 328.91*ln(Time-2004)þ1960
(2007,0.18),(2008,0.18),(2009,0.17),(2010,0.17),(2011,0.16),
(24) wind power generation capacity ¼ INTEG (“newly-added
(2012,0.15),(2013,0.15),(2014,0.14),(2015,0.14),(2016,0.13),
capacity,”1.264)
(2017,0.13),(2018,0.13),(2019,0.12),(2020,0.12),(2021,0.12),
(25) newly-added capacity ¼ increment of investment*10000/(con­
(2022,0.11),(2023,0.11),(2024,0.11),(2025,0.11),(2026,0.1),
struction and installation costs per capacity per year*20)
(2027,0.1),(2028,0.1),(2029,0.1),(2030,0.1),(2031,0.09),
(26) construction and installation costs per capacity ¼(10104*(Time-
(2032,0.09)))
2004)^(-0.148))*100/10
(55) interregional transmission capacity ¼ INTEG (newly-added
(27) benchmark price ¼ 0.004*(Time-2004)þ0.404
transmission capacity,0.01)
(28) ideal proportion of wind power generation ¼ IF THEN ELSE
(56) newly-added transmission capacity ¼ interregional transmission
(Time>2020, 0.08, 0.05)
capacity*increase rate of interregional transmission capacity
(29) the gap ¼ ideal proportion of gas power generation-actual pro­
(57) peak load capacity ¼ Thermal power installed capacity*0.01 þ
portion of wind power generation
hydropower installed capacity*0.05þGas-fired power installed
(30) policy ¼ IF THEN ELSE(gap between ideal and actual>0.05,2, IF
capacity*0.1
THEN ELSE(gap between ideal and actual>0.04, 1.8, IF THEN
(58) peak load factor ¼ IF THEN ELSE(peak load capacity>40, 0.6, IF
ELSE(gap between ideal and actual>0.03, 1.6, IF THEN ELSE(gap
THEN ELSE(peak load capacity>30, 0.55, IF THEN ELSE(peak
between ideal and actual>0.02, 1.4, IF THEN ELSE(gap between
load capacity>20, 0.5, IF THEN ELSE(peak load capacity>10,
ideal and actual>0.01, 1.2, 1) )) ))
0.45, 0.3))))
(31) revenue ¼ wind power generation*(benchmark price þ subsidy)
(59) wind power curtailment ¼ A FUNCTION OF(adjust factor,
*100/10000
transmission factor, For heating, For hydrogen production, peak
(32) subsidy ¼ A FUNCTION OF(policy)
load factor)
(33) net revenue ¼ revenue-cost
(34) maintenance cost ¼ coefficient*wind power generation
capacity*1000 References
(35) coefficient ¼ 0.0655*(Time-2004)þ2.9275
Angus, J., Raphael, J., Heffron, D.M., 2014. Rethinking the scope and necessity of energy
(36) other generation cost ¼ wind power generation*11*1000/10000 subsidies in the United Kingdom. Energy Res. Soc. Sci. 3, 1–4.
(37) total construction cost ¼ wind power generation capacity*con­ Aslani, A., Helo, P., Naaranoja, M., 2014. Role of renewable energy policies in energy
struction and installation costs per capacity per year dependency in Finland: system dynamics approach. Appl. Energy 113, 758–765.
Cherp, A., Vinichenko, V., Jewell, J., Suzuki, M., Antal, M., 2017. Comparing electricity
(38) cost¼(maintenance cost þ other generation cost þ total con­ transitions: a historical analysis of nuclear, wind and solar power in Germany and
struction cost)/10000 Japan. Energy Policy 101, 612–628.

14
X. Xu et al. Energy Policy xxx (xxxx) xxx

Faezipour, M., Ferreira, S., 2018. A system dynamics approach for sustainable water Qin, X., Fu, N., Zhong, S., 2018. Scenario Analysis of the Grid Parity in Various Wind
management in hospitals. IEEE Syst. J. 12, 1278–1285. Resource Areas in 2020. Wind Energy, pp. 70–73 (in Chinese).
Fendri, D., Chaabene, M., 2019. Hybrid Petri Net scheduling model of household Research Group of the Institute of Economics, Chinese Academy of Social Sciences, 2012.
appliances for optimal renewable energy dispatching. Sustain. Cities Soc. 45, China’s long-term growth path, efficiency and potential growth. Econ. Res. 11, 4–17.
151–158. Rosales-Asensio, E., Borge-Diez, D., Blanes-Peiro, J.-J., Perez-Hoyos, A., Comenar-
Feng, T., Yang, Y., Yang, Y., 2018. What will happen to the power supply structure and Santos, A., 2019. Review of wind energy technology and associated market and
CO2 emissions reduction when TGC meets CET in the electricity market in China? economic conditions in Spain. Renew. Sustain. Energy Rev. 101, 415–427.
Renew. Sustain. Energy Rev. 92, 121–132. https://doi.org/10.1016/j. Sharifian, A., Ghadi, M.J., Ghavidel, S., Li, L., Zhang, J.F., 2018. A new method based on
rser.2018.04.079. Type-2 fuzzy neural network for accurate wind power forecasting under uncertain
Forrester, J.W., 1969. Urban Dynamics. The MIT Press, Cambridge. data. Renew. Energy 120, 220–230.
Forrester, J.W., 1971. World Dynamics. Wright-Allen Press, Cambridge. Shukla, A., Singh, S.N., 2016. Clustering based unit commitment with wind power
Guo, Y., 2018. Analysis on the income for the grid parity of onshore Wind power. Wind uncertainty. Energy Convers. Manag. 111, 89–102.
Energy Ind. 36–46. State Grid Energy Research Institute Co, L., 2018. China’s New Energy Power Generation
Hache, E., Palle, A., 2019. Renewable energy source integration into power networks, Analysis Report in 2018. China Electric Power Press, Beijing, China.
research trends and policy implications: a bibliometric and research actors survey Wang, M., 2018. Research on China’s wind power and photovoltaic power generation
analysis. Energy Policy 124, 23–35. subsidy gap and large proportion of abandoned electricity. Int. Econ. Rev. 67–85þ
Hassan, Q.-U., 2013. Understanding the dynamics of electricity generation capacity in 66.
Canada: A system dynamics approach. Energy 59, 285–294. https://doi.org/ Wang, X., Zou, H., 2018. Study on the effect of wind power industry policy types on the
10.1016/j.energy.2013.07.029. innovation performance of different ownership enterprises: evidence from China.
He, Y.X., Jiao, J., Chen, R.J., Shu, H., 2018. The optimization of Chinese power grid Energy Policy 122, 241–252.
investment based on transmission and distribution tariff policy: a system dynamics Wang, F.Y., Wen, X.Y., Li, X.H., 2019. Study on dynamic impact of electricity price
approach. Energy Policy 113, 112–122. subsidy on renewable energy power generation. Price Theory Pract. 4, 54–58 (in
Hernandez, C.V., Gonzalez, J.S., Fernandez-Blanco, R., 2019. New method to assess the Chinese).
long-term role of wind energy generation in reduction of CO2 emissions - case study Ward, K., 2011. The world in 2050: quantifying the shift in the global economy. https
of the European Union. J. Clean. Prod. 207, 1099–1111. ://warwick.ac.uk/fac/soc/pais/research/researchcentres/csgr/green/foresigh
Hsu, C.-W., 2012. Using a system dynamics model to assess the effects of capital subsidies t/economy/2011_hsbc_the_world_in_2050_quantifying_the_shift_in_the_global_econo
and feed-in tariffs on solar PV installations. Appl. Energy 100, 205–217. my.pdf.
Huang, J., 2018. An empirical analysis of the impact of subsidy policy on the Wu, L., Li, H., 2017. Analysis of the development of the wind power industry in China -
development of wind power industry. Ecol. Econ. 34, 115–120. from the perspective of the financial support. Energy Sustain. Soc. 7.
Kedron, P., Bagchi-Sen, S., 2018. Explanations of firm survival in renewable energy in Wu, Y.N., Chen, K.F., Yang, Y.S., Feng, T.T., 2015. A system dynamics analysis of
the United States: a study of ethanol refineries. Eur. Plann. Stud. 26, 2237–2255. technology, cost and policy that affect the market competition of shale gas in China.
Khansari, N., Vesaghi, A., Mansouri, M., Mostashari, A., 2017. The multiagent analysis of Renew. Sustain. Energy Rev. 45, 235–243.
social progress in energy behavior: the system dynamics methodology. IEEE Syst. J. Xi, J., 2017. Multi-party game for grid parity of wind power. Energy 26–28 (in Chinese).
11, 2062–2071. Xia, F., Song, F., 2017. Evaluating the economic impact of wind power development on
Li, Q., 2017. There Is Still a Long Time for Wind Power to Reach the Grid Parity. China local economies in China. Energy Policy 110, 263–270.
Electric Power Enterprise Management, pp. 41–43 (in Chinese). Xiao, W.H., Ye, J., 2016. Comparison of the mechanism of renewable energy price and
Li, H., Yu, Y., Xie, Y.T., Zhang, J., 2016. CouldWind and PV energies achieve the grid policy choice. J. Jiangxi Univ. Financ. Econ. 1, 21–27 (in Chinese).
parity in China until 2020. Filomat 30, 4173–4189. Yan, X.H., Zhang, X., Gu, C.H., Li, F.R., 2018. Power to gas: addressing renewable
Lin, B., Omoju, O.E., Okonkvvo, J.U., 2016. Factors influencing renewable electricity curtailment by converting to hydrogen. Front. Energy 12, 560–568.
consumption in China. Renew. Sustain. Energy Rev. 55, 687–696. Yang, X.L., He, L.Y., Xia, Y.F., Chen, Y.F., 2019. Effect of government subsidies on
Liu, X., Ma, S., Tian, J., Jia, N., Li, G., 2015. A system dynamics approach to scenario renewable energy investments: the threshold effect. Energy Policy 132, 156–166.
analysis for urban passenger transport energy consumption and CO2 emissions: a Yao, X.L., Liu, Y., Qu, S.Y., 2015. When will wind energy achieve grid parity in China? -
case study of Beijing. Energy Policy 85, 253–270. connecting technological learning and climate finance. Appl. Energy 160, 697–704.
Marcella, N., Massimo, T., 2017. Are renewable energy subsidies effective? Evidence Yin, G.Z., Zhou, L., Duan, M.S., He, W.M., Zhang, P., 2018. Impacts of carbon pricing and
from Europe. Renew. Sustain. Energy Rev. 74, 412–423. https://doi.org/10.1016/j. renewable electricity subsidy on direct cost of electricity generation: a case study of
rser.2016.12.032. China’s provincial power sector. J. Clean. Prod. 205, 375–387.
National Energy Administration, N., 2017. Notice of the National Energy Administration Yin, X., Zhang, W., Zhao, X., 2019. Current status and future prospects of continuously
on the Publication of Demonstration Projects for the Grid Parity of Wind power[No. variable speed wind turbines: a systematic review. Mech. Syst. Signal Process. 120,
49, [2017]]. 326–340.
National Renewable Energy Center, Renewable Energy Development Center, E.R.I., Zendehboudi, A., Baseer, M.A., Saidur, R., 2018. Application of support vector machine
National Development and Reform Commission, 2018. China’s Renewable Energy models for forecasting solar and wind energy resources: a review. J. Clean. Prod.
Industry Development Report. China Economic Publishing House, Beijing, China. 199, 272–285.
Nguyen, C., Ma, C., Hailu, A., Chalak, M., 2016. Factors influencing calculation of Zhao, G.Y., 2019. Analysis of the on-line price of land-based winds from the cost of
capacity value of wind power: a case study of the Australian National Electricity electricity. Northwest Hydropower 7, 1–9 (in Chinese).
Market (NEM). Renew. Energy 90, 319–328. Zhao, E.D., Liu, P.K., 2013. Policy effects of feed in tariff and tradable green certificate
OECD, 2012. Looking to 2060: long-term global growth prospects. OECD Econ. Policy —based on the development of biomass power industry in China. Ind. Technol. Econ.
Pap. 3. 9, 125–137 (in Chinese).
Polaris Wind Power Network, P., January 2019. China’s Renewable Energy Subsidy Gap
Was100 Billion and CPPCC Members Recommended Reform Subsidy Model (in
Chinese).

15