Energy conservation, environmental and economic value of the wind power priority dispatch in China

Accepted Manuscript Energy conservation, environmental and economic value of the wind power priority dispatch in China Xiaoli Zhao, Suwei Liu, Fengguang Yan, Ziqian Yuan, Zhiwen Liu PII:

S0960-1481(17)30213-6

DOI:

10.1016/j.renene.2017.03.032

Reference:

RENE 8625

To appear in:

Renewable Energy

Received Date: 18 March 2016 Revised Date:

10 February 2017

Accepted Date: 12 March 2017

Please cite this article as: Zhao X, Liu S, Yan F, Yuan Z, Liu Z, Energy conservation, environmental and economic value of the wind power priority dispatch in China, Renewable Energy (2017), doi: 10.1016/ j.renene.2017.03.032. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Energy Conservation, Environmental and Economic Value of the

2

Wind Power Priority Dispatch in China

3

Xiaoli Zhao1∗, Suwei Liu2, Fengguang Yan2, Ziqian Yuan1, Zhiwen Liu2

4

1 School of Business Administration, China University of Petroleum-Beijing, Beijing, China

5

2 School of Economics and Management, North China Electric Power University, Beijing, China

6

Abstract: In recent years, the annual curtailment ratio of wind power in China has been

7

around 15%. Although the priority dispatch policy for wind power has been put in place

8

in the country, it is not well implemented. One of the reasons for this is that wind power

9

is viewed by some power system operators as “garbage power” due to its inherent

10

intermittency nature. The improvement of social awareness of the value of wind power

11

deployment is crucial for the effective implementation of the priority dispatch policy. In

12

this paper, we analyze the impact of the priority dispatch of wind power on energy

13

conservation and environmental improvement, and then compare the generation cost

14

between wind power and coal-fired power while considering environmental

15

externalities. We conclude that the energy intensity in the system decreases as the

16

proportion of wind power generation increases, and the energy intensity is minimized

17

when the proportion reaches 15%-20% provided a small amount of wind power

18

curtailment is allowed. Meanwhile, we confirm that the wind power priority dispatch

19

has significant positive impacts on environmental improvement. Further, we find that

20

wind power generation cost is very close to coal-fired power cost when environmental

21

cost is taken into account.

22

Keywords: Wind power priority dispatch; Environmental value; Energy conservation;

23

Economic cost; China

AC C

EP

TE D

M AN U

SC

RI PT

1

24 ∗

Corresponding Author Xiaoli Zhao, School of Business Administration, China University of

Petroleum-Beijing, E-mail: [email protected] 1

ACCEPTED MANUSCRIPT 1

1 introduction In recent years, China has been troubled by an extensive, severe, and long period of

3

haze. Since the beginning of winter 2015, there have been large regions of serious haze

4

in China’s northern areas. Beijing’s government has released two times of haze red

5

alerts, with the highest level of alert occurring in December 2015. During this time,

6

primary and middle school students cannot go to class, some enterprises cannot

7

produce, and people cannot conduct their usual travel schedules due to vehicle

8

restrictions. Although haze is caused by complicated factors, such as weather and air

9

pressure conditions, coal burning is one of the most significant causes. China’s energy

10

mix is dominated by coal. According to BP, coal consumption accounted for 66% of the

11

total primary energy in China in 2014, well above the worldwide average level of 28%;

12

moreover, China’s coal consumption accounted for 50% of worldwide consumption.

13

On the other hand, coal used for electricity generation in China is only approximately

14

50% of the total coal consumption; in some developed countries, coal used for

15

electricity generation is more than 95% of the total electricity generation. Hence, to

16

improve China’s environmental quality, two types of substitution are important: one is

17

the substitution of electricity for coal and the other is the substitution of renewable

18

energy for coal-fired electricity. Among various renewable energies, wind power

19

technology is the most mature; hence, wind power is the preferred renewable energy in

20

China.

AC C

EP

TE D

M AN U

SC

RI PT

2

21

While wind power in China has seen dramatic growth since the beginning of this

22

century, wind power curtailment has occurred since 2010. China’s wind power was 2

ACCEPTED MANUSCRIPT curtailed by 12300 GWh, 20800 GWh, and 16200 GWh in 2011, 2012 and 2013,

2

respectively [1-3]; accordingly, the curtailment ratio of wind power in these three years

3

was 16%, 20% and 10%, respectively. In 2014, the severe wind power curtailment was

4

mitigated by a curtailment ratio of 8%; this was because there was less wind than in

5

previous years. As of 2015, the wind power curtailment has become more serious. In

6

2015, wind power curtailment in China reached 15%1.

RI PT

1

The Energy Saving Power Generation Dispatch Policy (Trial) issued in 2007

8

stipulates that to promote renewable energy grid integration, wind power, solar power,

9

and other types of renewable energy that do not have any adjustment ability should

10

enjoy priority dispatched. Both the Renewable Energy Law and the Renewable Energy

11

Law (revised) promulgated in 2005 and 2009 respectively define the principle of wind

12

power priority dispatch. However, the high ratio of wind power curtailment implies that

13

the policy of wind power priority dispatch has not been implemented effectively in

14

China. There are many factors that are attributed to the difficulty of wind power priority

15

dispatch implementation, such as an inadequate grid infrastructure [4] [5], lack of

16

flexible power generation plants, a restricted power trade market, and so on [6]. To

17

overcome these obstacles, China must realize the value of wind power priority dispatch

18

and use this information to improve the mechanism of wind power priority dispatch.

19

Although some studies have shown that renewable energy development contributes

20

significantly to environmental improvement [7-8], currently, there is still controversy

AC C

EP

TE D

M AN U

SC

7

1

Data source: China National Energy Administration: The situation of wind power industry development in 2015. http://www.nea.gov.cn/2016-02/02/c_135066586. 3

ACCEPTED MANUSCRIPT surrounding the energy, environmental, and economic value of the wind power priority

2

dispatch/development [9-13]. Zhang et al. (2010) argue that the extension of wind

3

power generation in a power system will increase the total operating cost in certain

4

contexts [9]. Zhang et al. (2011) also illustrate that the power dispatch aimed at

5

minimizing wind power curtailment may cause thermal units to significantly deviate

6

from their optimal economic operating area, thereby increasing coal consumption in the

7

power system [10]. Elliston et al. (2016) evaluated the incremental costs of more

8

renewable energy supply and concluded that incremental costs are found to increase

9

approximately linearly as the renewable energy share grows from zero to 80% [11].

10

Contrary to these three studies, Milligan et al. (2009) stated that wind power integration

11

into the grid results in fuel savings and refuted the notion that wind’s variations would

12

increase system fuel consumption [12]. Meanwhile, Destek (2016) analyzed the

13

relationship between renewable energy consumption and economic growth in newly

14

industrialized countries. They argued that renewable energy development causes

15

positive or negative impact on GDP in different countries [13].

EP

TE D

M AN U

SC

RI PT

1

The controversy surrounding the impact of the wind power priority dispatch on

17

energy consumption and economic cost in a power system affects the understanding of

18

the wind power priority dispatch and the effective implementation of this policy, which

19

leads to randomness and volatility in the process of implementing the wind power

20

priority dispatch. The aim of this study is to analyze the energy conservation,

21

environmental value, and economic cost of a wind power priority dispatch in China. We

22

will focus on the following three issues: first, the impact of the wind power priority

AC C

16

4

ACCEPTED MANUSCRIPT dispatch on energy consumption in a power system; second, the environmental value of

2

the wind power priority dispatch; and third, an economic cost comparison between

3

wind power generation and coal-based power generation taking environmental

4

externalities into consideration.

RI PT

1

The rest of the paper proceeds as follows: Section 2 summarizes the current

6

literature; Section 3 introduces our research methodologies; and Section 4 discusses the

7

data that are used. Our results and analyses are presented in the fifth section. Section 6

8

concludes the paper with a summary of our main findings and a discussion of policy

9

implications.

2 Literature Review

M AN U

10

SC

5

Along with the rapid development of renewable energy, the optimal power

12

dispatch between renewable energy and fossil fuel energy to minimize energy

13

consumption/emission reduction/economic costs has recently become a hot research

14

topic. Using the “Mixed Integer Programming (MIP)” method to study western

15

Denmark, Göransson and Johnson (2009) explored the dispatch strategy that yields the

16

lowest system fuel consumption and cost [14]. They found that the start-up time and

17

related costs and minimum load level of the power generating units had a significant

18

impact on the results. Zhou et al. (2009) used a hybrid particle swarm and interior point

19

optimization algorithm with smoothing technology and considered the constraints of up

20

and down spinning reserves to analyze the impact factors of a dynamic economic

21

dispatch in a wind power integrated system [15]. They concluded that in the context of

22

full wind power utilization, the energy consumption in the system could be minimized

AC C

EP

TE D

11

5

ACCEPTED MANUSCRIPT 1

by optimizing the power output of thermal units. Meanwhile, the multi-objectives of expanding wind power deployment,

3

mitigating emission/energy consumption, and reducing economic costs is another hot

4

topic in research [16-19]. Wu et al. (2010) used a multi-objective differential evolution

5

(MODE) algorithm method to study the environmental/economic power dispatch (EED)

6

problem with wind power integration [16]. They argued that the multi-objectives of

7

increasing wind power deployment, energy conservation and economic cost savings

8

could be realized by optimizing the power dispatch model. Based on stochastic

9

programming, Long et al. (2011) proposed a short term optimization model that

10

integrates hydro-power and thermal power with wind power to analyze the

11

optimization of the dispatch strategy while considering energy conservation and

12

environmental protection [17]. They concluded that in the context of wind power

13

priority dispatch, the goal of reducing energy consumption and pollution emissions in a

14

power system can be realized. Zhao et al. (2013) used the GAMS simulation method

15

and the real data of the thermal power units, wind power and hydropower generation,

16

and load in China’s Liaoning province to conclude that progress in environmental

17

improvement could be achieved while simultaneously reducing economic losses

18

through a joint environmental and economic dispatch across thermal, wind, and

19

hydropower [18]. By stating that the allocation of system generation capacity is highly

20

dependent on the values of the reserve cost coefficient, penalty cost coefficient and

21

penalty emissions coefficient, Jin et al. (2014) treated the economic environmental

22

dispatch (EED) incorporating wind power as a multi-objective stochastic programming

AC C

EP

TE D

M AN U

SC

RI PT

2

6

ACCEPTED MANUSCRIPT 1

problem [19]. In summary, most current studies focused on the power dispatch optimization of

3

wind power, thermal power and hydro power; the impact of wind power integration on

4

the energy consumption/environment and economic cost in the power system has

5

largely been ignored. We have only found very limited studies that compare the

6

difference in environmental influence between wind power and coal-fired power

7

generation. Xue et al. (2015) compared CO2, SO2, NOX and PM10 emissions between

8

coal-fired plants and wind farms based on the life cycle method [20]. The results

9

showed that the substitution of wind power for coal-fired power per kWh can reduce the

10

emissions of CO2, SO2, NOx and PM10 by 97.48%, 80.38%, 57.31%, and 30.91%,

11

respectively. Chen et al. (2011) used a typical wind farm in China’s Guangxi province

12

as a case study to calculate the environmental effect of various stages in a wind farm life

13

cycle period [21]. They concluded that the emission of CO2 equivalent was 0.002 kg for

14

1 MJ of electricity by the wind farm plant, 108 times less than those of the average coal

15

plant in China. Also based on the life cycle method, Zhao and Wang (2014) analyzed

16

the difference in CO2 emissions between wind power and coal-fired power generation

17

[22], and argued that by substituting wind power for coal-fired power generation, CO2

18

could be reduced by 0.867 kg/ kWh; and taking the benefits of CO2 emissions reduction

19

into account, the cost of wind power generation was only 0.027 RMB Yuan/(kWh)2

20

higher than that of thermal power generation. To the best of our knowledge, this study is

21

the only one that explored the economic cost of wind power generation while

AC C

EP

TE D

M AN U

SC

RI PT

2

2

The conversion rate between RMB and US dollar was $1 = RMB6.94 on Jan. 1, 2017. 7

ACCEPTED MANUSCRIPT 1

considering environmental externalities. However, this study was only concerned about

2

the impact of the environmental cost of CO2, other types of pollutant emissions, such as

3

PM, SO2, and NOx, were not considered. The contribution of this study lies in two important aspects: first, we analyze the

5

impact of wind power priority dispatch on energy consumption in a power system and

6

attempt to answer the following two questions: Is wind power priority dispatch always

7

beneficial for energy conservation? If the answer is no, then what is the turning point at

8

which wind power curtailment can reduce the energy intensity in the power system?

9

Second, we compare the cost of wind power and coal-fired power generation to show

10

the economic value of the priority dispatch of wind power taking into the

11

environmental externalities resulted from CO2, dust (PM), SO2, and NOx.

12

3 Methodologies

13

3.1 Objective function

TE D

M AN U

SC

RI PT

4

Our objective function is to minimize the economic cost. The economic cost of a

15

power system consists of four parts: energy consumption from the operation of

16

coal-fired units, energy consumption from units starting and stopping for peak shaving

17

(the part inside square bracket), the punishment cost in the context of wind power

18

curtailment, and the reserve cost of coal-fired units.

19

AC C

EP

14

T

N

T

t=1

i=1

t=1

∑ ∑Ui,tOci (Pi,t ) +Ui,t (1−Ui,t−1)Sci,t +Ui,t−1(1−Ui,t )Sdi,t  +β∑(Pwt*, −Pwt, ) +fi (ri ) (1)

20

where T represents time, N is the number of coal-fired power units, and U it is the

21

operating state of unit i at moment t , which has two values, U it =0 represents the 8

ACCEPTED MANUSCRIPT stopped units and Uit =1 represents units that are in working condition. Pi t is the

2

coal-fired power generation for unit i at time t , and Oci ( Pi ,t ) is energy consumption

3

when coal-fired units are in operating status, which is generally expressed as a

4

quadratic function of the power of units [23-25], that is:

RI PT

1

5

Oci ( Pi,t ) = a1i Pi,2t + b1i Pi ,t + c1i

6

where a1i , b1i and c1i are generation cost parameters.

7

Sci ,t (in Eq. (1)) is the energy consumption for unit i that starts at time t . S di ,t is

8

the energy consumption for unit i that stops at time t . Pw*,t is wind power generation

9

that can be dispatched without curtailment at time t . Pw,t is the actual wind power

10

generation at time t . β is the priority coefficient, which is bigger than the maximum

11

value of the total coal consumption in the power system, incurring a steep cost when

12

even a small part of wind power is curtailed. fi (ri ) is the cost function of the reserve

13

capacity of coal-fired units; according to [23], it is:

SC

M AN U

TE D

14

(2)

f i ( ri ) = a2i + b2i rit + c2 i rit2

(3)

where a2i , b2i and c2i are reserve cost coefficients and rit represents the reserve

16

capacity of generation of unit i at time t . If a unit is not connected to the grid at time t ,

17

then the reserve capacity is at the upper limit of its output.

18

3.2 Constraint conditions

19

AC C

EP

15

(1) Constraint of Power balance N

20

∑U i =1

21

i ,t

Pi ,t + Pw,t = PL ,t

(4)

where N is the total number of coal-fired units in a system, Pw,t is the wind power

9

ACCEPTED MANUSCRIPT 1

generation at time t , and PL ,t is the load value at time t .

2

(2) Constraint of coal-fired power generation

3

U i ,t Pi ,min ≤ Pi ≤ U i ,t Pi ,max

6

7

respectively, for unit i . (3) Constraint of the coal-fired power ramping rate:  Pi ,t −1 − Pi ,t ≤ ∆Pi ,down   Pi ,t − Pi ,t −1 ≤ ∆Pi ,up

RI PT

5

where Pi ,min and Pi ,max are the minimum and maximum power generation,

SC

4

(5)

(6)

where ∆Pi , down and ∆Pi ,up are the down and up climbing rates of unit i .

9

(4) Technical constraints of a wind turbine

11 12 13

0 ≤ Pw,t ≤ Pw*,t

(7)

where Pw*,t is wind power generation that can be dispatched at time t ; Pw,t is actual wind power generation at time t .

TE D

10

M AN U

8

(5) Constraint of the spinning reserve of coal-fired plants Up-spinning reserve constraints:

15

Sui ,t = min( Pi tmax − Pi t , T10 × rui )

(8)

16

Pi tmax = min( Pi t −1 + rui × T60 )

(9)

AC C

17

EP

14

where S ui ,t is up-spinning reserve for unit i at time t within 10 minutes; rui is the

18

up-climbing rate for coal-fired unit i ; Pi tmax is the maximum power generation for unit

19

i at time t ; Pi t is power generation for unit i at time t ; T10 is the response time of

20

spinning reserve, which is 10 min; Pi t −1 is the power generation for unit i at time t -1 ;

21

and T60 is an operating period, which is 60 minutes.

22

Eq. (8) means that the additional power generated by coal-fired plants when wind 10

ACCEPTED MANUSCRIPT power can’t meet its promised generation (up-spinning reserve) should be constrained

2

by two conditions: one is the difference between the maximum power generation it can

3

reach at time t and the actual power generation at this time; the other is the ramp rate of

4

unit i for 10 minutes. For example, a unit can generate 600MW at time t ( Pi tmax = 600 ),

5

and its actual power generation at this time is 400MW ( Pi t = 400 ), then the possible

6

up-spinning reserve would be 200MW. However, coal-fired power generation increase

7

can’t be finished immediately, and it should be constrained by unit ramp-up rate (the

8

speed of adding power output). For example, if the ramp-up rate of unit i for 10

9

minutes is 100MW, then within 10 minutes, the up-spinning reserve for this unit is only

10

100MW instead of 200MW; but within 20 minutes, the up-spinning reserve for this unit

11

will be 200MW.

M AN U

SC

RI PT

1

Eq. (9) means that the maximum power generation at time t is constrained by the

13

power generation at time t − 1 and the additional power output in the next one hour. For

14

example, in the context of cold-start up of coal fired power plants (plants start up from a

15

status of fire off), no additional power will be generated in the next one hour since the

16

time of cold-start up is generally more than several hours. Then, the maximum power

17

generation at time t is the same as that at t − 1 .

19 20

EP

AC C

18

TE D

12

Down spinning reserve constraints: S di ,t = min( Pi t − Pi tmin , T10 × rdi )

(10)

Pi tmin = max( Pi min , Pi t −1 − rdi × T60 )

(11)

21

where S di ,t is down-spinning reserve for unit i at time t within 10 minutes; rdi is

22

down-climbing rate for unit i ; and Pi tmin is the minimum power generation for unit i at 11

ACCEPTED MANUSCRIPT 1

time t . Pi min is the minimum power output that coal-fired power unit i can attain. Similar with the meaning of Eq. (8) and (9), Eq. (10) means down spinning reserve

3

(the space of coal fired power generation can be reduced for integrating more wind

4

power generation) should also be constrained by two conditions: one is that a unit must

5

meet the minimum power generation. In general, the minimum power generation of a

6

coal fired unit is around 40% of its stored capacity. For example, the minimum power

7

generation of a 600MW unit is around 240MW. Hence, in Eq. (10), if the current power

8

generation for unit i is 400 MW ( Pi t = 400 ), its minimum power generation is 240MW

9

( Pi tmin = 240 ), then the possible down-spinning reserve would be 160MW. Meanwhile,

10

the down-spinning reserve is also constrained by ramp-down rate of a unit (the speed of

11

reducing power output), the situation is similar with that of Eq. (8). For example, if the

12

possible reducing power output for unit i is 160MW, the ramp-down rate of the unit is

13

100MW for 10 minutes, then the down spinning reserve will be 100MW instead of

14

160MW.

TE D

M AN U

SC

RI PT

2

Eq. (11) means the minimum power generation of a unit at time t ( Pi tmin ) depends

16

on the minimum power generation the unit can reach ( Pi min ) and the difference between

17

the actual power generation at time t − 1 and the reducing speed of power output within

18

one hour. For example, if the minimum power generation of a unit is 240MW

19

( Pi min =240 ), the actual power generation of the unit at time t − 1 is 400MW

20

( Pi t -1 = 400 ), the ramp-down rate of the unit is 200MW per hour ( rdi = 200 ), then the

21

minimum power generation of a unit at time t ( Pi tmin ) is 240 MW.

22

AC C

EP

15

(6) The minimum start/stop time constraint: 12

ACCEPTED MANUSCRIPT 1

on  (U i , t −1 − U i ,t )(Ti , t −1 − Ti , on ) ≥ 0  off  (U i , t − U i ,t −1 )(Ti , t −1 − Ti , off ) ≥ 0

2

where Ti ,ont −1 and Ti ,offt −1 is the continuous running time and stopping time for unit i at

3

time t -1 , respectively; Ti ,on and Ti ,off represents the minimum continuous running time

4

and minimum stopping time for unit i , respectively.

RI PT

(12)

Eq. (12) can be divided into two parts. The first part, (U i ,t −1 − U i ,t ) or (U i ,t − U i ,t -1 ) ,

6

represents the status of unit i . If t = 1 , then U i ,t −1 =U i ,0 , means the unit is in the status of

7

stopping at previous stage; and U i ,t =U i ,1 , means the unit is in the status of operating at

8

the current stage. (U i ,t −1 − U i ,t ) indicates that unit i starts up from stopping; and

9

(U i ,t − U i ,t -1 ) represents that unit i stops from starting up.

M AN U

SC

5

The second part of Eq. (12), (Ti ,ont −1 − Ti ,on ) or (Ti ,offt −1 − Ti ,off ) , represents the actual

11

continuous operating time of unit i must be more than the minimum permitted

12

continuous operating time, or the actual continuous stopping time of unit i must be

13

more than the minimum permitted continuous stopping time. For example, if Ti ,on =5 ,

14

means the minimum permitted continuous running time of unit i is 5 hours, in this

15

context, Ti ,ont −1 ≥ 5 , means the actual continuous running time of unit i must be more

16

than 5 hours. Similarly, if the minimum permitted continuous stopping time of unit i is

17

5 hours ( Ti ,off =5 ), the actual continuous stopping time of unit i must be more than 5

18

hours ( Ti ,offt −1 ≥ 5 ).

19

3.3 Reserve capacity model that considers uncertainty

AC C

EP

TE D

10

20

We construct a reserve capacity model that considers the uncertainty of wind

21

power and the power load based on the probability method. A probabilistic model of

22

reserve capacity can be used to obtain the up-regulated and down-regulated reserve 13

ACCEPTED MANUSCRIPT 1

capacity by using some confidence level with regard to the forecasting uncertainty of

2

the wind power and load to ensure safe operation.

4

The constraint of the up-spinning reserve capacity with an expected probability is as follows:

RI PT

3

N

5

Sust = ∑ Sui,t ≥ PLt × L% + Pw* t × wu %

(13)

i =1

where S ust is the total up-spinning reserve in a system at time t , S ui ,t is up-spinning

7

reserve for unit i at time t within 10 minutes, which is determined by Eq. (8); and PLt

8

is the anticipated power load; L % is the demand for up-spinning reserve caused by the

9

forecast error of total power load; Pw*

M AN U

SC

6

t

is the anticipated wind power generation; and

10

wu % is the demand for up-spinning reserves caused by the forecast error of wind power

11

generation.

Eq. (13) means the total up-spinning reserve should be no less than the sum of the

13

reserve required by the forecast error of total power load and the forecast error of wind

14

power generation. For example, the anticipated power load is 5GW at time t ( PLt = 5 ),

15

the forecast error of total power load is 5% ( L% = 5% ); the anticipated wind power

16

generation at time t is 1GW ( Pw*,t = 1 ), and the forecast error of wind power generation

17

is 10% ( wu % = 10% ), then the total up-spinning reserve should be no less than 0.35GW

18

(5×5%+1×10%=0.35).

19 20

21

AC C

EP

TE D

12

The constraint of down-spinning reserve capacity with expected probability is as follows: N

Nw

i =1

y =1

S dst = ∑ S di ,t ≥ (∑ Wymax +Pwt ) × wd %

(14)

14

ACCEPTED MANUSCRIPT where S dst is the total of down-spinning reserves in a system at time t ; S di ,t is the

2

total of down-spinning reserves for unit i at time t within 10 minutes, which is

3

determined by Eq. (10); Wymax is the maximum wind power generation for unit y ; Pwt is

4

the anticipated wind power generation at time t ; N w is the total number of wind units;

5

and wd % is the demand for down-spinning reserve caused by the forecast error of wind

6

power generation.

RI PT

1

Different from Eq. (13), the total of down-spinning reserves in Eq. (14) ( S ust )

8

should consider both the maximum of total wind power generation and the anticipated

9

power generation at time t . This is because down-spinning reserve is referred as the

10

power generation space that should be reduced by coal fired plants for extra wind power

11

integration out of the original planning. The extra wind power generation out of the

12

original planning depends on not only the anticipated power generation, but also the

13

maximum of total wind power generation. For example, if the maximum of total wind

TE D

M AN U

SC

7

Nw

14

power generation is 1GW ( ∑ Wymax =1 ), the anticipated wind power generation is

EP

y =1

0.8GW ( Pwt =0.8 ), the demand for down-spinning reserve caused by the forecast error of

16

wind power generation is 5% ( wd %=5% ), then the total of down-spinning reserves

17

should be no less than 0.09 GW [(1+0.8) ×5%=0.09].

18

AC C

15

With the expansion of the forecasting period, the forecast error of wind power

19

generation increases, thus increasing the demand of the up-spinning reserve capacity.

20

To describe the effect of the dynamic system dispatching reflected by the forecasting

21

period of wind power, wu % in Eq. (13) and wd % in Eq. (14) are used to follow the equal

22

growing ratio rule of different research periods. 15

ACCEPTED MANUSCRIPT 1

3.4 Life cycle method The life cycle method is used to calculate the environmental cost of coal-fired

3

plants and wind farms. The steps include: acquisition of production resources,

4

plant/farm construction, plant/farm operation, and plant/farm decommissioning. The

5

equation is as follows: n

n

m

i

i

j

QE = ∑ QEi = ∑∑ Ecij × QEf ij

(15)

SC

6

RI PT

2

where QE is the total CO2/pollutant emission in a life cycle, QEi is the

8

CO2/pollutant emission in stage i , Ecij is the consumption of material j in stage i ,

9

and QEf ij is the CO2/pollutant emission factor of material j in stage i .

11

Based on the CO2/pollutant emission, we can calculate the economic cost of environmental externalities using Eq. (16). n

12

C = ∑ QEm × Pm m =1

TE D

10

M AN U

7

(16)

where C is the total environmental externality cost, QEm is the emission of

14

CO2/pollutant m , Pm is the unit environmental cost of emission m , and n equals the

15

total types of emissions.

16

4 Data source

AC C

17

EP

13

The data regarding the parameters of coal-fired units has been collected from

18

Attaviriyanupap et al. (2002) [26]. The data used to calculate CO2 and pollutant

19

emissions for the coal-fired power output in a life cycle has been collected from Datang

20

Zhaluteqi Wind Farm Project I, with a 49.5MW capacity from 33 wind turbines located

21

in Inner Mongolia, China. The power load data have been collected from [27]. The data 16

ACCEPTED MANUSCRIPT used to calculate CO2 and pollutant emissions for wind power output in a life cycle have

2

been collected from Huaneng Jiutai Power Plant Project II, located in Jiutai city, Jilin

3

province, with an installed capacity of 2×660 MW of Super Critical Units. The wind

4

power generation data are shown in Fig. 1-3.

RI PT

1

Fig. 1 shows the wind power generation data for each hour in a month and the

6

variation characteristics; the wind power output is referred to as the percentage of the

7

average wind power generation in one month to the total installed capacity of wind

8

power. Fig. 2 and Fig. 3 show the wind power curve on a typical day in the summer and

9

winter, respectively.

M AN U

SC

5

Fig. 1 here.

11

Fig. 2 here.

12

Fig. 3 here.

13

For the spinning reserve data, according to Zhou et al. (2009) [28], the demand for

14

up-spinning ( wu % ) and down-spinning reserves ( wd % ) that is caused by the forecast

15

error of wind power generation is between 10%-30% and 30%-50%, respectively; the

16

spinning reserve is valued according to the equal proportion principle of the dispatch

17

period; the demand for up-spinning caused by the forecast error of the power load ( L % )

18

is assumed to be 5%. The coefficients of the reserve cost function are shown in table 1.

AC C

EP

TE D

10

19

Table 1 here.

20

The component and weight of a wind power turbine are based on the introduction

21

of the Dongqi FD77-1500 wind power turbine (Table 2), which is used in this study’s

22

wind farm case; it is commonly used turbine in China. The material data for the 17

ACCEPTED MANUSCRIPT 1

components of wind power turbines have been collected from [29] and are shown in

2

Table 3. Table 2 here.

4

Table 3 here.

5

The components and materials (mainly steel) used in building coal-fired units of

6

600 to 660MW are shown in Table 4, and the materials used for building the coal-fired

7

plant in this study are shown in Table 5. Table 4 here.

9

Table 5 here.

SC

M AN U

8

RI PT

3

5 Results and discussions

11

5.1 Impact of the wind power priority dispatch on energy consumption

TE D

10

Coal consumption in the power system of this study continuously decreases as

13

wind power generation increases (Table 6). The decrease is primarily caused by energy

14

conservation during the process of power generation (operating energy consumption).

15

Contrary to the continuous decrease of operating energy consumption, the reserve

16

energy consumption increases as wind power generation increases (Table 6). This

17

happens because as wind power generation increases, more reserve capacity is needed;

18

in this context, more coal-fired units should be operated at a low output level, which

19

leads to an increase in energy intensity. Meanwhile, as wind power generation increases,

20

start-stop energy consumption first increases and then decreases. This is because with

21

more wind power generation, more coal-fired power units are needed to provide start

AC C

EP

12

18

ACCEPTED MANUSCRIPT 1

and stop peak shaving services. However, when wind power generation increases to a

2

certain level, in this study, 30% of the total power generated (Table 6), some coal-fired

3

power units cannot be restarted once they are shut down. Table 6 here.

5

In the power system, start-stop energy consumption and reserve energy

6

consumption are relatively small; total energy consumption depends on the operating

7

energy consumption. Table 6 shows that as wind power generation increases, total

8

energy consumption significantly decreases. Fig. 4 shows the change in energy

9

intensity (the energy consumption for average hourly power generation) that results

10

from an increase in wind power generation. It shows that with an increase in wind

11

power generation, energy intensity 3 continuously decreases, from 323.11 gram (g) of

12

standard coal equivalent (sce)/kWh to 265.37 g sce /kWh, as the percentage of wind

13

power generation increases from 5% to 40%.

TE D

M AN U

SC

RI PT

4

Fig. 4 here.

15

Fig. 4 shows that an increase in wind power generation leads to a decrease in

16

energy intensity in the power system. However, we don’t know if to minimize the

17

energy intensity, wind power should be curtailed? This question is answered in Table 7.

18

It shows that when the proportion of wind power generation is comparatively small,

19

wind power should be dispatched first to minimize energy consumption (energy

20

intensity). As the proportion of wind power generation increases, the system cannot

AC C

EP

14

3

It is defined as the quantity of standard coal consumption of per unit power generation. 19

ACCEPTED MANUSCRIPT accommodate all of the wind power. Therefore, the minimization of energy

2

consumption occurs with the curtailment of wind power. In this study, curtailment of

3

wind power begins between 15% and 20%. However, at this point, the wind power

4

curtailment ratio is very low. For example, the wind power curtailment ratio is only

5

0.21% when the wind power generation proportion is 20% (Table 7).

7

Table 7 here.

SC

6

RI PT

1

5.2 Environmental value of the wind power priority dispatch

We first calculate the emissions of wind farms from a life cycle perspective. Based

9

on the data in Table 2 and Table 3 and according to the studies from Wang and Zhang

10

(2012) [30], Ruan et al. (2010) [31], Zhang et al. (2008) [32], Ye et al. (2007) [33], and

11

Li et al. (2009) [34] about the environmental externalities of the materials used to

12

manufacture wind turbines, we can derive the CO2 and pollutant emissions during the

13

manufacture of wind turbines (Table 8). Meanwhile, we can derive the emissions in the

14

transportation stage of the wind turbine from the data in Table 2 and Table 3 as well as

15

based on the reference [22], [35-36] (Table 9). Reference [22] states that, in general, the

16

transportation distance of wind turbines is 2500 km by railway and 400 km by road; [35]

17

states that 30% and 70% of the components of wind turbines are transported by railway

18

and highway, respectively; and [36] provides the emission intensity during the process

19

of materials transportation both by railway and highway.

AC C

EP

TE D

M AN U

8

20

Table 8 here.

21

Table 9 here.

20

ACCEPTED MANUSCRIPT The primary materials used in wind farm construction include those consumed in

2

the infrastructure construction, wind turbine installation, and electrical equipment

3

installation. According to the study results of [31], [32], and [37-40], which are related

4

to the environmental externalities of the relative materials in a life cycle period, and the

5

materials quantities used in the process of wind farm construction calculated by [22],

6

we can obtain the emissions in the stage of wind farm construction (Table 10).

RI PT

1

Table 10 here.

8

Furthermore, we need to calculate the emissions of a wind farm during its

9

operation and decommissioning stages. The emissions in the operation stage are caused

10

by both the replacement and maintenance of damaged components and the

11

transportation of the related components and workers in the wind farm. According to

12

[22], the life span of a wind farm is typically 20 years; the diesel oil consumed in this

13

stage is 1495.44 L. Guo (2012) argues that in the entire life span of a wind farm, a blade

14

and 15% of the total components will need to be replaced [41]. Moreover, according to

15

the study results of [29], which pertain to the weight of the blade and the material used,

16

and the study results of [32] and [34], which pertain to the environmental externalities

17

of the related materials in a life cycle, we can derive the emissions in the wind farm

18

operation stage (Table 11). The emissions in the wind farm decommissioning stage

19

account for an estimated 10% of the emissions in the wind turbine manufacturing stage

20

[35]; the results are shown in Table 12.

21

AC C

EP

TE D

M AN U

SC

7

Table 11 here.

21

ACCEPTED MANUSCRIPT Table 12 here.

2

Based on Tables 8 through 12, we can derive the total emissions of CO2 and

3

pollutants for a life cycle period of the wind farm in this study. The wind power

4

generated from the wind farm is 117.57 GWh per year, and the life cycle of a wind farm

5

is, on average, 20 years. Hence, we can calculate the emissions for an average unit of

6

wind power generation from the wind farm, which is shown in Table 13.

SC

RI PT

1

Table 13 here.

8

The wind farm in this study, the Datang Zhaluteqi Wind Farm Project I, is a

9

representative example of wind farms in China. On one hand, the capacity of this farm

10

is 49.5 MW. According to the EPISC (2013), of the 205 wind farms built in 2013, 175

11

are installed with a capacity ranging from 47.5 MW to 50 MW. On the other hand, the

12

wind turbines (including fans and towers) used in this farm are the Dong Fang

13

FD77-1500, which is a common wind turbine that is used in China. Hence, the results

14

presented in Table 13 represent the general emission levels of China’s wind farms.

EP

TE D

M AN U

7

Second, we calculate the emissions of coal-fired plants from a life cycle

16

perspective. Di et al. (2005) calculate the emission intensity in the coal mining and

17

transportation processes [42]. Because we know the coal consumption of the Huaneng

18

Jiutai Power Plant Project II is 4.213×106 tons, we can calculates the emissions in the

19

coal production and transportation stages (Table 14).

AC C

15

20

Table 14 here.

21

The main equipment used in coal-fired power plant construction include the boiler, 22

ACCEPTED MANUSCRIPT turbine and generator; the material used to produce this equipment is primarily steel,

2

and the steel needed to produce this equipment can be calculated using the 600-660MW

3

generator manual of Shanghai. Moreover, according to [30], which pertains to the

4

environmental externalities of the steel production process, we can derive the emissions

5

in the main equipment manufacturing stage (Table 14).

RI PT

1

To maintain consistency with the analysis for wind power farm emissions, we

7

calculate the emissions in the equipment transportation process assuming that 30% and

8

70% of the equipment of coal-fired power plant are transported by railway and highway,

9

respectively. According to the China Statistical Yearbook (2014), in 2013, the average

10

transport distances via railway and highway were 735 km and 181 km, respectively,

11

which are taken as the transport distance of the equipment. Based on [36], which

12

pertains to the emission intensity of railway transport and highway transport, we can

13

obtain the emissions in the main equipment transportation stage (Table 14).

M AN U

TE D

14

SC

6

The emissions in the coal-fired power plant construction stage include the emissions caused by the installation of the equipment

16

and the construction of the main buildings. The materials used in the equipment

17

installation and main building construction processes are calculated according to

18

Organization & Design Standard on the Power Factory Construction of 2×660MW and

19

Construction Standard of Coal fired Power Plants of 2×600MW, respectively.

20

According to [30], [31], [39], and [40], which calculate the environmental externalities

21

of the production of related materials, we can obtain the emissions from the coal-fired

22

power plant construction stage (Table 14).

boiler, turbine, and generator

AC C

EP

15

23

ACCEPTED MANUSCRIPT The emissions in the coal-fired plant operation stage include the emissions caused

2

by burning coal and waste disposal. Di et al. (2005) calculate the emission intensity of

3

burning coal in coal-fired power plants [42]; the desulfuration, denitration, and

4

dedusting in coal-fired power production is given as 90%, 82.3%, and 99.7%,

5

respectively [42]. According to the Huaneng Jiutai Power Plant Project II Environment

6

Impact Report (simplified edition), the annual output of desulfurization gypsum and ash

7

is 48.6 thousand tons and 410.4 thousand tons, respectively; the ash yard distance is 4.5

8

km. Based on the above information, we can calculate the emissions in the coal-fired

9

plant operation stage (Table 14).

M AN U

SC

RI PT

1

The Huaneng Jiutai Power Plant Project II has an installed capacity of 2*660 MW

11

of Super Critical Units. According to the China’s Electric Power Industry Statistical

12

Compilation (EPISC) (2013), there are 7223 coal-fired power units, the capacities of

13

which are not less than 6 MW; only 522 units have capacities that are over 600 MW, or

14

7.2% of the total capacity. This suggests that this project represents a relatively high

15

level of technology in China. To generalize our research, when evaluating the emissions

16

per unit of coal-fired power generation, we use the national average coal consumption

17

rate (coal consumption per unit of power generation) of 302 g sce/kWh in 2013 instead

18

of the coal consumption rate of this specific plant (282.4 g sce/kWh).

EP

AC C

19

TE D

10

As to the emissions in the coal-fired power plant decommissioning stage,

20

according to [43], the emissions in this stage are 10% of the emissions in the coal-fired

21

plant construction stage. Hence, we can calculate the emissions in this stage (Table 14).

22

While the coal-fired power plant in this study has a high technological ranking, we 24

ACCEPTED MANUSCRIPT 1

substitute the national average coal consumption rate level of coal-fired plants for the

2

consumption of this plant to ensure that our study is representative of the national level

3

in Table 14. Third, comparing the emissions of wind power generation (Table 13) with the

5

emissions of coal-fired power generation (Table 14), we argue that there is a great

6

environmental value to be obtained by substituting wind power for coal-fired power.

7

Specifically, substituting 1 kWh of coal-fired power with wind power reduces SO2,

8

NOx, CO2, and PM by 0.7556 g/kWh, 0.721 g/kWh, 791.92 g/kWh, and 0.4591 g/kWh,

9

respectively. Our SO2, NOx, CO2 reduction results are similar to those of Zou and Ma

10

(2003) and Gao et al. (2012). Zou and Ma (2003) and Gao et al. (2012) do not calculate

11

the PM reduction. According to the report issued by the China Electricity Council, wind

12

power generation in 2014 of 156.3 TWh accounted for 2.8% of the total power

13

generation; the thermal power generation (dominated by coal-fired power generation)

14

as 4173.1 TWh, accounted for 75.2% of the total power generation in 20144. If wind

15

power generation increases by 1% and thermal power generation decreases accordingly

16

by 1% in China, SO2, NOx, CO2, and PM will decrease by 4.19×105, 4.0×105, 4.4×108,

17

and 2.5×105 tons, respectively.

18

5.3 Economic feasibility of the wind power priority dispatch considering

19

environmental cost

AC C

EP

TE D

M AN U

SC

RI PT

4

4

Data source: Analysis on China’s electricity industry operation conditions in 2014. China

Electricity Council, 2015, 02, 03. Last access: 2016, 01, 30. http://www.jcoal.com/haomei/dianli/news/html/1502/20150203_312318.html 25

ACCEPTED MANUSCRIPT 1

Jiang et al. (2013) evaluates the environmental cost of SO2

NOx CO2, and PM

emissions [44]; however, the above study uses the 2010 level of purchasing power. To

3

accurately reflect the current environmental cost of emissions, we recalculate the

4

economic cost of the above emissions using the 2014 purchasing power level; the

5

results are shown in Table 15.

RI PT

2

Table 15 here.

7

Based on Tables 13, 14, and 15 as well as Eq. (17), we can obtain the

8

environmental cost of wind power generation and coal-fired power generation; they are

9

0.003 RMB Yuan/kWh and 0.208 RMB Yuan/kWh, respectively. n

ECW = ∑ EQwm × Pm 10

m =1 n

ECC = ∑ EQcm × Pm

(17)

TE D

m =1

M AN U

SC

6

where ECw and ECc are the environmental cost of wind power and coal-fired

12

power, respectively; EQwm and EQcm are pollutant m emission for wind power

13

generation and coal-fired power (per kWh), respectively; Pm is the environmental cost

14

of pollutant m .

AC C

15

EP

11

According to [22], the economic cost of coal-fired power generation and wind

16

power generation are 0.307 RMB Yuan/kWh and 0.44 RMB Yuan/kWh, respectively.

17

He et al. (2013) [45] argue that the cost of ancillary services for wind power generation,

18

which include the peak shift cost, frequency modulation cost, and spinning reserve cost

19

[46], is 0.084 RMB Yuan/kWh. Hence, the economic cost for wind power generation is

20

0.524 RMB Yuan/kWh including the cost of ancillary services. 26

ACCEPTED MANUSCRIPT Next, we can derive the comprehensive cost of wind power and coal-fired power

2

by considering both economic environmental costs, which are 0.527 RMB Yuan/kWh

3

and 0.515 RMB Yuan/kWh, respectively. The results indicate that the wind power

4

generation cost is approximately equal to the coal-fired power cost when the

5

environmental cost is taken into consideration, meaning that implementing a wind

6

power priority dispatch, even from an economic aspect, is feasible.

7

6 Concluding remarks

SC

RI PT

1

China’s power mix has long been overwhelmingly dominated by coal-fired power,

9

the cost of which is lower than in the U.S., the European Union, India, and much lower

10

than the domestic gas-fired power cost. Scaling up the deployment of renewable energy

11

is critical for the country to address its deteriorating environmental pollution, and to

12

coordinate its economic growth and energy consumption as well. In this study, we

13

explore the energy conservation, environmental, and economic value of wind power

14

priority dispatch in China. China’s wind power generation ratio to its total power

15

generation is still around 3%; however, its average wind power curtailment is more than

16

10% annually. The consciousness and attitudes regarding a wind power priority

17

dispatch (or incentive policies of wind power generation) play an important role in the

18

implementation of these policies and the reduction of the wind power curtailment ratio.

19

However, the controversy surrounding the value of wind power priority dispatch still

20

exists in China; some scholars (and policy makers) question the energy conservation

21

and economic value of wind power priority dispatch. This study attempts to address the

22

current doubts casting on the value of wind power priority dispatch in China.

AC C

EP

TE D

M AN U

8

27

ACCEPTED MANUSCRIPT Our first conclusion is that the energy intensity of the power system in this study

2

decreases as the proportion of wind power generation increases; specifically, the energy

3

intensity decreases from 323.11 g sce/kWh to 265.37 g sce/kWh, when wind power

4

generation increases from 5% to 40% in this case study. To realize the minimization of

5

the energy intensity in this study, wind power curtailment should occur when wind

6

power generation is between 15% and 20% of the total power generated. However, the

7

wind power curtailment ratio is as low as 0.21% when the wind power generation ratio

8

is 20%. Even when the wind power generation ratio reaches 40% of the total power

9

output, the wind power curtailment ratio that minimizes energy intensity is 8.30%,

10

which is lower than the current average national level. Therefore, we conclude that to

11

improve energy conservation, China’s wind power priority dispatch policy needs to be

12

effectively implemented.

TE D

M AN U

SC

RI PT

1

Our second conclusion is that wind power priority dispatch has a significant value

14

on environmental improvement. We have shown that by substituting 1 kWh of wind

15

power for 1 kWh of coal-fired power, the emission of SO2, NOx, CO2, and PM will be

16

reduced by 0.7556 g/kWh, 0.721 g/kWh, 791.92 g/kWh, and 0.4591 g/kWh,

17

respectively.

AC C

18

EP

13

Our final conclusion is that wind power priority dispatch has economic feasibility.

19

When taking external environmental costs into account, the costs of wind power and

20

coal-fired power generation are 0.527 RMB Yuan/kWh and 0.515 RMB Yuan/kWh,

21

respectively; the economic generation costs of wind power and coal-fired power are

22

0.524 RMB Yuan/kWh (including an ancillary service cost of 0.084 RMB Yuan/kWh) 28

ACCEPTED MANUSCRIPT and 0.307 RMB Yuan/kWh, respectively; the environmental costs of wind power and

2

coal-fired power are 0.003 RMB Yuan/kWh and 0.208 RMB Yuan/kWh, respectively.

3

The overall cost of wind power is very close to that of coal-fired power. The

4

technological progress in wind power generation combined with increasing concern

5

about the environment will lead to increased value of wind power. Hence, the economic

6

value of wind power priority dispatch is expected to become more apparent in the near

7

future.

SC

RI PT

1

This paper makes three main contributions. First, we empirically investigate the

9

impact of wind power priority dispatch on energy intensity, stating that while wind

10

power priority dispatch will not always ensure energy intensity minimization, an

11

increase in wind power generation will lead to continuous decreases in energy intensity.

12

Second, we evaluate the environmental value of substituting wind power for coal-fired

13

power by examining four types of emissions: CO2, SO2, NOx, and PM (to our

14

knowledge, previous studies have only explored the emissions of CO2, SO2, and NOx).

15

Third, we calculate the overall cost of wind power generation and coal-fired power

16

generation, considering the economic cost and environmental cost as well as the

17

ancillary service cost of wind power generation.

TE D

EP

AC C

18

M AN U

8

An important policy implication from this analysis is that China needs to

19

strengthen the implementation of its current wind power priority dispatch policy. We

20

have shown that wind power generation has a significant impact on energy

21

conservation and environmental improvement. However, the relatively positive

22

environmental externalities of wind power cannot be reflected by the current power 29

ACCEPTED MANUSCRIPT market in China, hence, the mandatory wind power priority dispatch policy should be

2

implemented at present. Along with the mitigation of market failure caused by

3

environmental externalities (for example, with the improvement in carbon tax or carbon

4

emission trading mechanism), it is advisable to implement market approaches to

5

promote wind power generation through competition [47-48]. Mignon and Bergek

6

(2016) also argued that incentive policy is an important trigger for emerging investors'

7

decisions in building renewable energy projects [49]. China has taken a series of

8

incentive policies to promote wind power and other kinds of renewable energy

9

generation [50]. Nevertheless, there is a long way to go for China to promote large scale

SC

M AN U

10

RI PT

1

increase of wind power and other renewable energy by effective incentive policies. It should be noted that our conclusions regarding the impact of wind power priority

12

dispatch on energy intensity should be viewed with caution. Despite our attempts to

13

make our study representative of the current power market in China, the power system

14

in our study is simulated and may differ from that of the actual system.

15

Acknowledgments

EP

TE D

11

We would like to thank the National Natural Science Foundation of China (Grant

17

No. 71373078, 71573273) and the National Social Science Foundation of China (Grant

18

No. 14BJY063) for partially funding this study as well as the fund from China

19

University of Petroleum-Beijing (Grant No. 2462014YJRC052). We appreciate

20

professor Sufang Zhang’s help in correcting this paper.

21

References:

AC C

16

30

ACCEPTED MANUSCRIPT [1] LI junfeng ,etc. 2012 China Wind Energy Development Report. China Resources

2

Comprehensive Utilization of Renewable Energy Commission, China Renewable

3

Energy Society wind energy professional committee, the Global Wind Energy

4

Council, 2013.

5 6

RI PT

1

[2] Global wind power market development report 2014 [R] Global Wind Energy Council, 2015.

[3] Li junfeng. China Wind Energy Development Report in 2011.[R]China Resources

8

Comprehensive Utilization of Renewable Energy Commission, China Renewable

9

Energy Society wind energy professional committee, the Global Wind Energy

12 13 14

M AN U

11

Council, 2012.

[4] Li, J., 2010. Decarbonising power generation in China — is the answer blowing in the wind? Renewable and Sustainable Energy Reviews 14, 1154–1171.

TE D

10

SC

7

[5] Liao, C.P., Jochem, E., Zhang, Y., Farid, R.N., 2010. Wind power development and policies in China. Renewable Energy 35, 1879–1886. [6] Zhao, X. L., Zhang, S. F., Yang, R., Wang, M. Constraints on the effective

16

utilization of wind power in China: an illustration from the Northeast China Grid.

18 19

AC C

17

EP

15

Renewable and Sustainable Energy Reviewers, 2012, 16 (1), 4508-4514.

[7] Dogan, E., Seker, F. Determinants of CO2 emissions in the European Union: The role of renewable and non-renewable energy. Renewable Energy, 2016, 94, 429-439.

20

[8] Kaldellis, J.K., Apostolou, D., Kapsali, M., Kondili, E. Environmental and social

21

footprint of offshore wind energy. Comparison with onshore counterpart. Renewable

22

Energy, 2016, 92, 543-556. 31

ACCEPTED MANUSCRIPT 1

[9] Zhang Q., Li H., Gao S. Research on the impact of wind power on peak load

2

regulation and its rational utilization mode. Southern Power System Technology,

3

2010, 4(6), 18-22. (In Chinese) [10] Zhang, L. Z., Zhou, N., Wang, N. Economic comparison for different generation

5

scheduling with large scale wind power connected power system. Automation of

6

Electric Power Systems, 2011, 35(22), 105-110.

RI PT

4

[11] Elliston, B., Riesz, J., MacGill, I. What cost for more renewables? The incremental

8

cost of renewable generation-An Australian National Electricity Market case study.

9

Renewable Energy, 2016, 95, 127-139.

12

M AN U

11

[12] Milligan M, Porter K, DeMeo E, et al. Wind power myths debunked?! IEEE Power & Energy Magazine, 2009, 7(6), 89-99.

[13] Destek, A. M. Renewable energy consumption and economic growth in newly

TE D

10

SC

7

13

industrialized countries: Evidence from asymmetric causality test. Renewable

14

Energy, 2016, 95, 478-484.

EP

16

[14] Göransson, L., Johnsson, F. Dispatch modeling of a regional power generation system--Integrating wind power. Renewable Energy. 2009, 34: 1040-1049.

AC C

15

17

[15]Zhou wei, Peng li, Sun hui, et al. Dynamic economic dispatch in power system

18

including wind farms.Chinese Journal of Electrical Engineering, 2009, 29(25), 13-

19

18.

20

[16] Wu, L. H., Wang, Y. N., Yuan, X. F., Zhou, S.W. Environmental/economic power

21

dispatch (EED) problem using multi-objective differential evolution algorithm.

22

Electric Power Systems Research, 2010, 80, 1171-1181. 32

ACCEPTED MANUSCRIPT 1

[17] Long jun, Mo qunfang, Zeng jian. Power system energy saving optimization

2

scheduling strategy based on stochastic programming with wind farm. Grid

3

technology, 2011, 35(9), 133-138. [18] Zhao, X. L., Wu, L. L., Zhang, S. F. Joint environmental and economic power

RI PT

4 5

dispatch considering wind power integration: Empirical analysis from Liaoning

6

Province of China. Renewable Energy, 2013, 52, 260-265.

SC

8

[19] Jin, J. L., Zhou, D. Q., Zhou, P., Miao, Z. Environmental/economic power dispatch with wind power. Renewable Energy, 2014, 71, 234-242.

M AN U

7

9

[20] Xue, B., Ma, Z.X., Geng, Y., Heck, P., et al. A life cycle co-benefits assessment of

10

wind power in China. Renewable and Sustainable Energy Reviews, 2015, 41,

11

338-346.

[21] Chen, G. Q., Yang, Q., Zhao, Y. H. Renewability of wind power in China: A case

13

study of nonrenewable energy cost and greenhouse gas emission by a plant in

14

Guangxi. 2011, 15, 2322-2329.

17 18 19 20

EP

16

[22] Zhao, X. L., Wang, S. H. Economic evaluation of wind power generation based on benefits of carbon dioxide emission reduction. Electric Power, 2014, 47(8),

AC C

15

TE D

12

154-160. (In Chinese)

[23] Wood, W. G. Spinning reserve constrained static and dynamic economic dispatch. IEEE Transactions on Power Apparatus and Systems 1982, PAS -101, 381 -388.

[24] Wu, L. H, Wang YN, Yuan XF, Zhou SW. Environmental/economic power

21

dispatch (EED) problem using multi-objective differential evolution algorithm.

22

Electric Power Systems Research 2010, 80, 1171-1181. 33

ACCEPTED MANUSCRIPT 1

[25] Kuo, C. C. Generation dispatch under large penetration of wind energy

2

considering emission and economy. Energy Conversion and Management 2010, 51:

3

89-97. [26] Attaviriyanupap P, Kita H, Tanaka E, et al. A hybrid EP and SQP for dynamic

5

economic dispatch with non-smooth fuel cost function. IEEE Transactions on

6

Power Systems, 2002, 17(2), 411-416.

9 10 11 12

SC

Dalian University of Technology, 2010. (In Chinese)

M AN U

8

[27] Zhou W. Study on dynamic economic dispatch in power system studies.

[28] Zhou, W., Peng, L., Sun, H. Wei, Q. H. Dynamic economic dispatch in wind power integrated system. Proceedings of the CSEE, 2009, 29(25), 13-18. (In Chinese) [29] Martnez E, Sanz F, Pellegrini S, et al. Life cycle assessment of a multi-megawatt wind turbine. Renewable Energy, 2009, 34: 667- 673.

TE D

7

RI PT

4

[30] Wang, L. F., Zhang, L. S. Life cycle assessment of environmental impacts for the

14

whole steel production process. China Population, Resources and Environment,

15

2012, 22(11), 239-244. (In Chinese)

17 18

[31] Ruan, R. M., Zhong, S. P., Wang, D. Z. Life cycle assessment of two copper

AC C

16

EP

13

metallurgical processes: Bio-heap leaching and flotation—flash smelting. Multi Purpose Utilization of Mineral Resources, 2010, 3, 33-37. (In Chinese)

19

[32] Zhang, Q., Gong, X. Z., Wang, Z. H. Wang, W. et al. Life cycle assessment of

20

typical glass production line. Compiles of the Seminar on China National Glass

21

and Furnace Technology, 2008, 9.

22

[33] Ye, H. L., Ma, W. H., Yang, B., Ren, J. Y. et al. LCA study of metallurgical silicon 34

ACCEPTED MANUSCRIPT 1

process. Light Metals, 2007, 11, 46-49. (In Chinese) [34] Li, M., Wang, Z., Sun, D. Z. Study on life cycle assessment of polythene

3

production. Environmental Science & Technology, 2009, 32(5), 191-195

4

Chinese)

(In

RI PT

2

[35] Gao, C. K., Dong, J. H., Zhu, W. G., Wang, W. Environmental load analysis of

6

wind turbines based on life cycle assessment. Journal of N ortheastern U

7

niversity(N atural Science), 2012, 33(7), 1034-1037. (In Chinese)

9

[36] Gao, Y. X. Study of External Cost of Coal Fired Power Generation in China,

M AN U

8

SC

5

Dissertation, South China University of Technology, 2012. (In Chinese) [37] Chen, W. Q., Wan, H. Y., Wu, J. N., Shi, L. Life cycle assessment of aluminium

11

and the environmental impacts of aluminium industry. Light Metals, 2009, 5,

12

3-10. (In Chinese)

TE D

10

[38] Xu, X. N., Chen, Y., Zhang, S. S., Zhang, Y. et al. Life cycle assessment of

14

composite portland cement. Acta Scientiae Circumstantiae, 2013, 33(9),

15

2632-2638. (In Chinese)

17

[39] Zhou, X. X., Song, Z. L. The life cycle assessment of two kinds of concretes.

AC C

16

EP

13

Environmental Engineering, 2009, 27(S1), 472-475. (In Chinese)

18

[40] Hu, Z. Y., Tan, P. Q.,Lou, D. M. Dong, R. Q. Life cycle emission assessment of

19

disel and its alternative fuels. Chinese Internal Combustion Engine Engineering,

20 21 22

2007, 28(3), 80-84. (In Chinese) [41] Guo, M. X. Life cycle assessment of CO2 emissions of wind power, PV power and biomass power. Dissertation, Tsinghua University, 2012. (In Chinese) 35

ACCEPTED MANUSCRIPT 1

[42] Di, X. H., Nie, Z. R., Zuo, T. Y. Life cycle emission inventories for the fuels

2

consumed by thermal power in China China Environmental Science, 2005, 25(5),

3

632-635

(In Chinese)

[43] Liu, J. R., Qian, Y., Li, X. X. Life cycle assessment of coal-fired power generation

5

and its alternatives. Journal of China Coal Society, 2009, 34(1), 133-138

6

Chinese)

RI PT

4

(In

[44] Jiang, H., Zhao, J. Z., Zhao, X. L., Fan, C. Y. Quantitative analysis on

8

environmental externality of Chinese thermal power industry. Electric Power,

9

2013, 46(7), 126-132. (In Chinese)

M AN U

SC

7

[45] He, Y., Hu, J. F., Yan, Z. T., Shang, J. C. Compensation mechanism for ancillary

11

service cost of grid-integration of large-scale wind farms. Power System

12

Technology, 2013, 37(12), 3552-3557. (In Chinese)

TE D

10

[46] Bai, X. Z., Wang, G. Q., Wang, J. X., Chen, C. Influences of wind power

14

connecting on ancillary services of Northwest China Grid. Electric Power, 2009

15

(12), 73-76. (In Chinese)

17 18

[47] Zhao, X. L., Zhao, Y., Zeng, S. X., Zhang, S. F. Corporate behavior and

AC C

16

EP

13

competitiveness: Impact of environmental regulation on Chinese firms, Journal of Cleaner Production

2015, 86(1), 311-322.

19

[48] Zhao, X. L., Yin, H. T., Zhao, Y. Impact of Environmental Regulations on the

20

Efficiency and CO2 Emissions of Power Plants in China. Applied Energy, 2015,

21

149(1), 238-247.

22

[49] Mignon, I., Bergek, A. Investments in renewable electricity production: The 36

ACCEPTED MANUSCRIPT 1

importance of policy revisited. Renewable Energy, 2016, 88, 307-316. [50] Zhao, Z. Y., Chen, Y. L., Chang, R. D. How to stimulate renewable energy power

3

generation effectively? —China's incentive approaches and lessons. Renewable

4

Energy, 2016, 92, 147-156.

AC C

EP

TE D

M AN U

SC

RI PT

2

37

ACCEPTED MANUSCRIPT Tables

Parameters

Value

a2i

0.0074

b2i

51.62

c2i

117.73

RI PT

Table 1 Coefficient value in the reserve cost function

M AN U

SC

Sources: Authors’ calculation based on the data collected from field research in Liaoning province.

Table 2 Component and weight of a wind power turbine Component

Blade

Tower

Wind wheel

Cabin

Gear

Generator

Hub

Spindle

Weight（t）

6×3

90.4

15.9

5.6

16

6.8

16

7.15

TE D

Source: http://www.doc88.com/p-971466175230.html

Table 3 Materials for the manufacture of wind power turbines

150.95

AC C

Demand（t）

Steel

EP

Materials

Copper

Glass fiber

Silicon

Resin

3.5

20

0.3

1.1

Source：Martnez E, Sanz F, Pellegrini S, et al. Life cycle assessment of a multi-megawatt wind turbine. Renewable Energy, 2009, 34, 667- 673.

ACCEPTED MANUSCRIPT

Table 4 Materials used for building the coal-fired unit in this study

Steel consumption（t）

Boiler

Turbine

Generator

509.42

1052.52

490

Total

RI PT

Main equipment

2051.94

Source: http://www.docin.com/p-335996903.html；http://www.docin.com/p-450639808.html；

M AN U

SC

600-660MW generator manual of Shanghai.

Table 5 Materials used for building the coal-fired plant in this study Material

Concrete

Net

Steel（t） Reinforced

used

（m3）

frame（t）

Quantity

69406

248

2749

7925

5610

CO2（t）

19572.49

545.6

6047.8

17435

370.26

TE D

（t）

AC C

EP

Source: http://www.doc88.com/p-9522934644806.html

Aerated Concrete

Total

Masonry（m3） — 43971.15

ACCEPTED MANUSCRIPT

Table 6 Coal consumption in a power system (tce: tons of standard coal equivalent) operation energy

start-stop

Reserve

consumption

consumption

energy

energy

RI PT

θ

Total

consumption

consumption 25.7492

5%

440.2524

412.4960

2.0072

10%

401.8394

372.3124

2.3394

15%

381.9736

349.8306

2.5566

29.5864

20%

355.6202

320.9476

2.7976

31.8750

25%

339.2586

302.3238

2.8514

34.0832

30%

290.3952

287.4126

2.9824

36.5624

35%

276.3548

273.6326

2.7222

38.0028

40%

246.5516

259.5952

2.6562

41.2500

SC

M AN U

TE D

27.1874

Note: θ is the proportion of wind power generation to the total power generation in the

EP

system.

AC C

Table 7 Wind power curtailment ratio for minimizing energy consumption Wind power generation ratio to the total power generation ( θ )

Parameter

Wind power curtailment ration ( ε )

θ

5%

10%

15%

20%

25%

30%

35%

40%

ε

0

0

0

0.21%

1.23%

3.79%

5.87%

8.30%

Wind power curtailment ratio: the ratio of wind power curtailment to the possible maximum wind power generation.

ACCEPTED MANUSCRIPT

Table 8 Emissions in the manufacturing stage of wind turbines (33 turbines in all) NOx

CO2

18.711

40.524

12181.521

RI PT

PM 8.712

SC

Emissions（t）

SO2

SO2 0.561

NOx

CO2

PM

0.462

108.570

0.330

TE D

Emissions（t）

M AN U

Table 9 Emissions in the transportation stage of wind turbines (33 turbines in all)

Table 10 Materials used and relative emissions in the stage of wind farm construction

EP

Materials Used

Emissions(t)

Quantity of using（t）

SO2

NOx

CO2

PM

Steel

1911.8

3.84

11.03

3941.9

0.74

AC C

Material

Copper

8.9

0.28

0.25

97.09

Aluminum

78.16

2.68

1.1

669.52

0.07

Cement

2749.2

1.02

5.28

1660.52

29.47

Concrete

192

0.06

0.28

117

6.74

Diesel

26.87

0.143

0.419

103.01

0.061

8.02

18.36

6589.04

37.08

Total

ACCEPTED MANUSCRIPT

Table 11 Emissions in the stage of wind farm operation and maintenance NOx

CO2

4.13

4.8

1272.12

RI PT

PM 2.16

SC

Emissions（t）

SO2

SO2 1.87

NOx

CO2

PM

4.05

1218.15

0.87

TE D

Emissions（t）

M AN U

Table 12 emissions in the stage of wind farm decommissioning

Table 13 Emissions of wind power generation (g/kWh) Wind turbine

Wind turbine

EP

Emiss-

Wind farm

Wind farm

Wind farm Total

manufacture

transportation

construction

operation

decom.

SO2

8×10-3

3×10-4

3.4×10-3

1.8×10-3

8×10-4

1.42×10-2

AC C

ions

NOx

1.72×10-2

2×10-4

7.8×10-3

2×10-3

1.7×10-3

2.9×10-2

CO2

5.18

4.62×10-2

2.8

0.54

0.52

9.09

PM

3.7×10-3

1×10-4

1.58×10-2

9×10-4

4×10-4

2.09×10-2

ACCEPTED MANUSCRIPT

Table 14 Emissions of coal-fired power generation（g/kWh） Equipment

Power

Plant decommissioning

Coal mining

Total manufacture

construction

generation

RI PT

Emission

Plant

0.43

0.02

5.90×10-4

0.32

5.90×10-5

0.77

NOx

0.71

0.04

2.26×10-3

0.01

2.27×10-4

0.75

CO2

44.86

2.27

8.91×10-1

752.90

8.91×10-2

801.01

PM

0.19

0.01

4.02×10-2

0.24

4.02×10-3

0.48

M AN U

SC

SO2

Note: Coal mining includes coal production and transportation.

Emissions Environmental cost

AC C

EP

(RMB Yuan/kg)

TE D

Table 15 Environmental cost of emissions (RMB Yuan/kg) SO2

NOx

CO2

PM

34

22

0.18

24

ACCEPTED MANUSCRIPT Figures

RI PT

Wind power generation

SC

Month

TE D

Wind power generation

M AN U

Fig. 1 Average monthly wind power generation1

EP

Hour

1

AC C

Fig. 2 Wind power generation curve on a typical day in summer

The data in Fig 1 to Fig. 3 were collected from Texas in the United States for the following two

reasons: first, all of the wind power generation data in China were from after wind power curtailment, and we have no method to locate the original wind power generation data. Second, the wind power resource in Texas is the richest in the United States, which is similar to the status of the wind power resource in the Chinese province in this study.

Wind power generation

SC

Hour

RI PT

ACCEPTED MANUSCRIPT

EP

TE D

M AN U

Fig 3 Wind power generation curve on a typical day in winter

AC C

Fig. 4 Energy intensity change with a wind power generation increase

ACCEPTED MANUSCRIPT

Highlights

Wind power increase can promote energy intensity decrease significantly.




Wind power curtailment is very low for realizing energy intensity minimization.




Wind power priority dispatch has a big value on environmental improvement.




Wind power priority dispatch has economic feasibility.

AC C

EP

TE D

M AN U

SC

RI PT