Analysis of Energy Efficiency for Coal-fired Power Units Based on Data Envelopment Analysis Model

Available online at www.sciencedirect.com

ScienceDirect Energy Procedia 61 (2014) 904 – 909

The 6th International Conference on Applied Energy – ICAE2014

Analysis of Energy Efficiency for Coal-fired Power Units based on Data Envelopment Analysis Model Chenxi Song, Mingjia Li, Fan Zhang, Yaling He, Wenquan Tao Key Laboratory of Thermo-Fluid Science and Engineering, Ministry of Education, Xi’an Jiaotong University, Xian Ning West road, Xi’an, Shaanxi 710049, PR China

Abstract In this article, the non-parametric data envelopment analysis method˄DEA˅is employed to evaluate energy efficiency (EE) of coal-fired power units in China. The data set contains inputs and outputs of 34 coal-fired power units with the capacity of 600MW. Input-oriented CCR ˄Charnes-Cooper-Rhodes˅is employed. The value of CCR efficiency calculated in this paper is based on two input parameters: fuel consumption and auxiliary power consumption. The relations between EE and factors including, parameter of main steam, cooling method, capacity utilization rate and annual utilization hours are analysed. Some relation curves are fitted. © 2014 2014 The The Authors. Authors.Published Publishedby byElsevier ElsevierLtd. Ltd.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and/or peer-review under responsibility of ICAE Peer-review under responsibility of the Organizing Committee of ICAE2014 Keywords: Coal-fired power unit; data envelopment analysis (DEA); energy efficiency (EE)

1. Introduction The rising of the global temperature due to the greenhouse gas, especially carbon dioxide (CO2) emission is threatening the life of human being. Statistics show that energy consumption accounts for over 80% of the global anthropogenic CO2 emission [1], and traditional fossil energy resources like natural gas, oil and coal take a large proportion of total energy use in many countries. Power industry consumes considerable amount of fossil fuels. China is the largest power generator which accounts for 21.9% of world total power generation [2]. Statistics by China Electricity Council indicate that coal-fired thermal power plants takes 65.8% in total installed capacity in China [3]. Therefore, improving energy efficiency (EE) of coal-fired power units can contribute a lot for the ease of carbon emission in China. The DEA approach has been employed to analyse EE of power units by many researchers [4-11]. Among them, total factor EE evaluation which labour is included as an input was conducted by many researchers [4-6, 8, 9]. To analyse influencing factors of EE, regression method[4, 5, 8, 12] rather than the DEA method [7] is widely used. In this paper, EE analysis is conducted especially for coal-fired power units. The input and output parameters selected are only related to energy flow. From the respect of energy utilization, DEA models are used to analyse influencing factors of EE. Although the DEA method can not exclude the influence of other factors when analysing one factor, it

Corresponding author. Tel.: +86-298-266-9106; fax: +86-298-266-9106. E-mail address: [email protected].

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Organizing Committee of ICAE2014 doi:10.1016/j.egypro.2014.11.992

Chenxi Song et al. / Energy Procedia 61 (2014) 904 – 909

is effective to find predominant influencing factors. What’s more, results obtained in this paper can help to verify the feasibility of seeking predominant influencing factors of EE based on DEA models. 1.1. Model description In China, the most commonly used EE indicator of power units is called Standard Coal Consumption per Unit Power Supplied. However, this indicator can only depict overall EE of power units without showing any details. Efficiency evaluation based on data envelopment analysis (DEA) model is one of the methods which can help to solve this problem. In this paper, the input-oriented CCR ˄Charnes-Cooper-Rhodes˅model is used to calculate EE [13, 14]. The input-oriented version is adopted because the amount of electricity generated by all power plants is arranged and dispatched by state-owned group companies rather than determined by power plants. The software of DEAP (version 2.1) developed by Coelli is employed for solving equations of DEA models. 1.2. Selection of input and output parameters From the economic aspect, managers of coal-fired power plants concern a lot about coal and electricity consumption. The indicator of special EE here is defined as EE value based on DEA models with coal and electricity as two input parameters. Input and output parameters in our research are shown in Table1. Table 1. Selection of input and output parameters Input / Output

Parameter

Unit

Definition

Input 1

Coal consumption

105kgce

Coal consumption by boiler during operational stage

Input 2

Auxiliary electricity consumption

106kWh

Electricity consumption by auxiliaries for power units

Output 1

Electricity generated

106kWh

Total electricity generated by power units

Output 2

Capacity utilization rate

%

Utilization of the rated capacity of power units

2. Data description The data in our analysis are took from the public notification data of EE Competition of Fossil Power Units [15]. Thirty-four coal-fired power units with the rated capacity of 600MW are selected in this article. Different power units are identified with serial numbers. According to Raab and Lichty [16], the minimum number of decision making units (DMUs) should be greater than three times the number of inputs and outputs. This requirement is satisfied since here 34>3(2+2). Original data is list in the table of Appendix A. Statistical period for the data is a year. 3. EE analysis and discussion The general production flow chart of coal-fired power units is demonstrated in Fig. 1. The dotted lines demonstrate the statistical boundary. The right boundary lines marked with 1 and 2 indicate different statistical range. The indicator of Standard Coal Consumption per Unit Power Supplied is defined as the ratio of net electricity supplied over coal consumption. Right statistical boundary of this indicator is demarcated by the line marked with “1” (see Fig. 1). This is an effective indicator to evaluate the overall EE of power units. However, it treats the production process as a completely black box. We do not know which part affect EE of a power unit. The authors try to solve this problem by going a little deeper into the inner production process. As shown in Fig. 1, the right boundary is moved from “1” to “2” in our research. Several factors may affect EE of power units. In the following, affection of four factors including parameter of main steam, cooling method, capacity utilization rate and annual utilization hours on EE of power units are analysed.

905

906

Chenxi Song et al. / Energy Procedia 61 (2014) 904 – 909

Fig. 1. Schematic flow chart for electricity generation

3.1. Parameter of main steam The pressure of main steam is an important parameter for a power unit. The 34 power units are divided into two groups: supercritical power units and subcritical power units. Calculation results show that the average values of EE based on CCR model are 0.959 and 0.928 for supercritical power units and subcritical power units respectively. It indicates that increasing main steam pressure can increase EE of power units. Detailed EE distribution of power units in these two groups is demonstrated in Fig. 2. It can be found that some power units in the group of supercritical power units have lower EE than units in the group of subcritical power units. For example, power units “11” and “12” are such units. The authors analyse these two power units and found two reasons. First, condensers of these two power units are cooled by air directly. This is the type of cooling with the lowest efficiency. Second, capacity utilization rates of these two power units are 68.8 % and 69.79%, which are lower than other power units in the data set. Low capacity utilization rate can also lead to low EE. These two factors will be discussed later. Therefore, it can be said that the pressure of main steam has a positive affection on special EE based on CCR efficiency. Some power units which do not obey this rule are mainly due to the affection of other disadvantageous factors. 3.2. Cooling method The affection of cooling types on EE is shown in Fig. 3. There are four cooling types for condensers, namely water cooling with closed circulation (marked as “BS”), water cooling with open circulation (marked as “KS”), indirect air cooling (marked as “JK”) and direct air cooling (marked as “ZK”). The influence of cooling types is not obvious since there are uneven numbers of power units in each group. However, we can still find that power units cooled by water have larger EE compared with those cooled by air. In Fig. 3, the symbol of square represents the statistical average value, and the horizontal lines in the four bars represent the locations of 25%, 50% and 75% percentiles.

Fig. 2. EE comparison between supercritical and subcritical power units

Fig. 3. Statistical diagram of EE distribution for power units with different cooling styles

3.3. Capacity utilization rate and annual utilization hours The affection of capacity utilization rate and annual utilization hours on EE is demonstrated in Fig. 4 and Fig. 5, respectively. These two factors are normalized by the maximum value and minimum value among the 34 power units.

907

Chenxi Song et al. / Energy Procedia 61 (2014) 904 – 909

Relationships between the special EE and these two factors are obvious except for power units “29” and “30”. These two units are special due to two reasons. First, they are supercritical power units. Second, they are located in Liaoning province, where the temperature is relatively low in China. Low temperature in the cold side is a favourable factor for EE improvement in the Rankine cycle on which the electricity production is based. What’s more, condensers of these two units are cooled by water with open circulation. All these factors are favourable for EE. Therefore, these two power units are excluded when fitting the lines between EE and these two factors in Fig. 4 and Fig. 5. It can be found that EE and these two factors are positive related. The slope of the fitted line in Fig. 4 is larger than that in Fig.5, which indicates that the increment in capacity utilization rate will lead to larger increment of CCR efficiency compared with annual utilization hours.

1.00

29

30

y=0.120*x+0.863 R2=0.538

0.98

0.98

0.96

CCR efficiency

CCR efficiency

y=0.092*x+0.875 R2=0.336

1.00

0.94 0.92 0.90 0.88

0.96 0.94 0.92 0.90 0.88

0.86 0.0

0.2

0.4

0.6

0.8

1.0

Capacity utilization rate (non-dimensional) Fig. 4. The relation between EE and capacity utilization rate

0.86 0.0

0.2

0.4

0.6

0.8

1.0

Annual utilization hours (non-dimensional)

Fig. 5. The relation between EE and annual utilization hours

4. Conclusion We analyse EE of coal-fired power units based on CCR efficiency of DEA method in this paper. Compared with the conventional indicator of Coal Consumption per Unit Power generated, this indicator can show the EE of inner production process, which is helpful for the EE improvement. From the analysis of the relation between EE and five factors, it can be concluded that: 1) Increasing main steam pressure can help to improve EE of power units. 2) Water cooling is an efficient cooling style for power units. 3) The slope value of capacity utilization rate has larger influence on EE compared with annual utilization hours. Based on the conclusion, our advances for the design and operation of coal-fired power plants are as follows. 1) Higher parameters of steam should be selected as long as the material of power units can bear it. 2) Water cooling should be selected preferentially if the natural condition allows. Indirect air cooling should be selected rather than direct air cooling in water-deficient area. 3) High capacity utilization rate and annual utilization hours are very beneficial for EE of power units. To obtain high annul utilization hours, safety and reliability of power units should be enhanced by power plants. To obtain high capacity utilization rate, reasonable power dispatching should be conducted by Chinese grid companies. What’s more, since the effect of capacity utilization rate is larger than utilization hours, power units with low EE can be shut down for a while to maintain high capacity utilization rate of power units with high EE. Acknowledgements This work is supported by China National Energy Conservation Centre.

908

Chenxi Song et al. / Energy Procedia 61 (2014) 904 – 909

References [1] Quadrelli R, Peterson S. The energy-climate challenge: recent trends in CO2 emissions from fuel combustion. Energy Policy. 2007;35:5938-52. [2] British Petroleum. Statistical review of world energy. British Petroleum Company, London2013. [3] China Electricity Council. Lists of basis statistical data for power industry in 2011. Available at: ; 2013. [4] Lam P-L, Shiu A. A data envelopment analysis of the efficiency of China's thermal power generation. Utilities Policy. 2001;10:75-83. [5] Lam P-L, Shiu A. Efficiency and productivity of China's thermal power generation. Utilities Policy. 2004;24:73-93. [6] Fleishman R, Alexander R, Bretschneider S, Popp D. Does regulation stimulate productivity? The effect of air quality policies on the efficiency of US power plants. Energy Policy. 2009;37:4574-82. [7] Sarica K, Or I. Efficiency assessment of Turkish power plants using data envelopment analysis. Energy. 2007;32:1484-99. [8] Fallahi A, Ebrahimi R, Ghaderi SF. Measuring efficiency and productivity change in power electric generation management companies by using data envelopment analysis: A case study. Energy. 2011;36:6398-405. [9] Sueyoshi T, Goto M, Ueno T. Performance analysis of US coal-fired power plants by measuring three DEA efficiencies. Energy Policy. 2010;38:1675-88. [10] Liu CH, Lin SJ, Lewis C. Evaluation of thermal power plant operational performance in Taiwan by data envelopment analysis. Energy Policy. 2010;38:1049-58. [11] Shrivastava N, Sharma S, Chauhan K. Efficiency assessment and benchmarking of thermal power plants in India. Energy Policy. 2012;40:159-76. [12] Shanmugam KR, Kulshreshtha P. Efficiency analysis of coal-bassed thermal power generation in India during post-perform era. International Journal of Global Energy Issues. 2005;23:15-28. [13] Charnes A, Cooper WW, Rhodes E. Measring the efficieny of decision making units. European Journal of Operational Research. 1978;2:429-44. [14] Ramanathan R. An introduction to data envelopment analysis: A tool for performance measurement: Sage publications, New Delhi; 2003. [15] China Electricity Council. The notification about public announcement for data of power units with reated capacity of 600MW in national energy efficiency benchmarking and competition in 2012. Available at: ; 2013. [16] Raab R, Lichty RW. Identifying subareas that comprise a greater metropolitan area: the criterion of country relative efficiency. Journal of Regional Science. 2002;42:579-94.

First Author Biography Chenxi Song is a PhD candidate of Prof. Wenquan Tao in the department of Thermal Engineeringˈ Xi’an Jiaotong University, and a researcher in the Key Laboratory of Thermo-fluid Science and Engineering, Ministry of Education. Her Research mainly focuses on energy efficiency evaluation in industry. Corresponding Author Biography Wen-Quan Tao is currently a professor of Xi’an Jiaotong University and a member of Chinese Academy of Sciences. He is an associate editor of International Journal of Heat & Mass Transfer, International Communications in Heat & Mass Transfer, and ASME Journal of Heat Transfer. His research mainly focuses on heat transfer enhancement theory, mechanisms and engineering applications.

Appendix A. Original data for energy efficiency calculation in this paper Table A.1.Original data for energy efficiency calculation Power

Electricity

Capacity

Coal consumption

Oil consumption

Fresh water

Electricity

909

Chenxi Song et al. / Energy Procedia 61 (2014) 904 – 909

units

generated 6

utilization

consumption 4

3

consumption

10 kWh

%

10 t

t

m

106kWh

1

3382.41

71.88

100.37

13.86

70015.89

180.27

2

3722.86

73.78

112.04

11.45

77063.20

199.47

3

3745.89

77.21

111.18

0.01

8990.14

178.46

4

3872.97

74.97

114.74

0.01

9295.13

191.33

5

4069.00

77.50

121.96

48.00

83414.50

226.00

6

3899.00

74.89

117.37

65.00

81879.00

194.00

7

2906.36

78.50

89.69

183.00

6684.62

232.54

8

3636.32

79.69

113.04

171.38

8363.53

284.88

9

3252.15

76.74

96.82

80.20

66018.65

182.31

10

3819.53

77.73

111.05

19.50

77536.46

193.75

11

3087.24

68.80

98.33

434.02

8644.26

204.50

12

3174.63

69.79

101.11

280.77

8888.97

194.51

13

2906.26

71.52

92.86

91.88

11334.40

162.27

14

3074.31

70.11

98.19

86.62

11989.81

169.93

15

3438.43

71.57

108.29

601.16

6189.18

281.63

16

2496.48

68.76

80.19

1400.20

4493.67

209.09

17

3209.40

71.03

100.12

0.47

9949.14

254.29

18

3040.70

68.92

94.98

1.07

9426.17

241.71

19

3660.85

75.20

114.11

289.70

9884.28

296.69

20

3101.29

73.83

96.44

283.34

8373.49

254.88

21

3186.83

71.02

99.38

136.62

10516.55

250.29

22

3594.09

70.31

109.38

76.85

11860.49

275.78

23

3460.70

76.17

100.79

58.00

69214.00

180.70

24

3790.74

75.01

111.09

203.00

75814.80

199.67

25

3124.25

78.69

93.34

286.76

65609.25

165.68

26

3784.30

80.84

113.98

185.39

77578.15

199.94

27

3578.63

79.77

101.94

154.12

66204.66

190.20

28

3805.45

82.36

107.63

159.20

76109.00

199.74

29

2235.59

62.75

64.76

72.05

7601.01

99.41

30

2769.82

66.16

79.81

57.78

9417.39

115.87

31

3513.38

75.21

106.45

325.63

14053.51

168.24

32

3333.51

75.32

101.51

270.56

13334.02

148.46

33

3651.30

74.91

111.86

84.64

14605.20

180.64

34

3418.77

74.01

105.70

49.51

13675.08

159.03