Does change of industrial structure affect energy consumption structure: A study based on the perspective of energy grade calculation

Abstract

In this article, the grades of different kinds of energy sources are distinguished. Thus, we put forward an equivalent electric calculation method, which is compliant with the calculation of various energy resources that have different grades. Based on this aspect, we empirically analyzed the influence of industrial structure changing on energy consumption structure by analyzing panel data in 30 provinces of China from 2003 to 2013. Results showed that the calculated results of equivalent electric calculation method were more accurate because it considered the difference in grades between various energy sources. Industrial structure changing had a significant impact on energy consumption structure. The upgrading and rationalization of the industrial structure had a significant promotion on energy structure cleaning. In addition, technological progress was conducive to the clean development of energy structure, the decrease in energy price boosted energy structure cleaning, and the impact of economic level on energy consumption structure was not significant.

Keywords

Equivalent electric calculation method energy structure cleaning industrial structure upgrading industrial structure rationalization change of industrial structure

Introduction

As the driving force of economic production activities, energy has been taken as an important productive factor of the economic manufacture system by many scholars (Song and Zhou, 2015; Zhang and Choi, 2013; Zhang et al., 2013). With the rapid development of the economy, the consumption of energy resources in China is increasing rapidly at the same time, and fossil fuels are still the main types of energy consumption in this country, which also caused serious environmental pollution (Zhou et al., 2013a, 2013b). Therefore, the environment has become an important constraint for the sustained economic development in China. At present, “actively developing clean energy and optimizing the structure of energy consumption” are not only an important approach which is concerned by different sectors of society and can solve the contradiction between economic development and environmental protection but also the main policy guidance which is initiated by the Chinese government to develop the green economy. Therefore, it is of great significance to analyze the evolution of China’s energy consumption structure from the perspective of industrial structure change.

Statistical calculation of energy consumption

At present, there are two main statistical calculation methods for energy consumption in China: one is coal equivalent calculation method, this method counts other energy sources’ consumption by converting the average coal-fired power consumption of the current year into standard coal consumption; the other is calorific value calculation method which converts other energy sources into standard coal consumption according to their electric power equivalents. There are huge differences between the two methods in the statistical calculation of new energy sources such as hydropower, nuclear power and so on. When calculating the Chinese regional final energy consumption and the carbon emission of energy consumption, Li and Hu (2012), Zhou et al. (2013a, 2013b), Sun et al. (2015), and Guo et al. (2016) used energy converting into standard coal as the reference coefficient, which came from “Chinese Statistical Yearbook”. These reference coefficients used coal equivalent calculation method and calorific value calculation method as the two energy conversion methods, respectively, resulting in the discrepancy of final statistical calculation of economic consumption. In addition, the two calculation methods only considered the “quantity” of economic consumption, but ignored the different “quality” between different kinds of energies (Jang et al., 2006). For example, if we use calorific value calculation method to convert the 1 MJ calorific power of natural gas and coal into standard consumption, the results are same. In practical production, however, the average generated energy of 1 MJ calorific power of the above two energies are 0.55 MJ and 0.34 MJ, respectively. Obviously, there are significant differences in the qualities of various energy sources. In this article, we called this difference as “energy grade”. Ignoring the difference in various “energy grades” may cause deviations in the statistical calculation of energy consumption, which is not conducive to the optimal allocation of energy resources.

Industrial structure changing and energy consumption structure

In the earlier research literature, the industrial structure changing only refers to the industrial structure upgrading, it refers to the changing process of the industrial structure in which the dominant industry shifts from the primary industry to the secondary industry and then continuous to turn to the tertiary industry. When selecting upgrading indicator of the industrial structure, Zhang and Zhang (2015) believe that the decreasing trend of the added value primary industries in gross domestic product (GDP) in the time series is an important indicator of the industrial structure upgrading. Therefore, they suggested selecting the ratio of the added value of the secondary and tertiary industries to the total GDP value of the current year as the measurement index. But on the other hand, Duarte and Restuccia (2010) and Gan et al. (2011) believe that the proportion of the added value of the tertiary industry represents the characteristics of the second stage of industrial structure changing, and this proportion is more suitable for calculating the upgrading of Chinese industrial structure. Nowadays, the academia has a unanimous conclusion about the influence of the industrial structure upgrading on energy consumption structure, which states that the industrial structure upgrading is an important factor which can influence the regional energy structure change (Liu et al., 2011). The greater the proportion of the added value of tertiary industry (service industry) in GDP of the year, the greater the proportion of clean energy consumption.

As research continued, some scholars believe that the industrial structure upgrading cannot represent the entire content of the industrial structure changing. Gan and Zheng (2009), Peng et al. (2013), and Jiao (2015) all believe that industrial structure changing can optimize resource allocation and promote the development of industrial structure rationalization, and then propelled the industrial structure changes from low grade to high grade. Thus, industrial structure changing includes two dimensions: the industrial structure upgrading and the industrial structure rationalization. Industrial structure upgrading means the degree of coordination between different industries and the coupling of production system input and output structure. Gan and Zheng (2009) use “Theil index” to measure industrial structure rationalization, and it is the prevailing method at present. There is little empirical research about the effect of industrial structure rationalization on energy consumption structure which is an important content that we study in this topic.

Nowadays, studies about this topic have two limitations at home and abroad. On the one hand, the statistical calculation of energy consumption which academia used at present ignores various energy grades, so it cannot accurately calculate the consumption of energy conversion, thereby causing a bias in the measurement of energy consumption structure. On the other hand, industrial structure changing includes two aspects: the industrial structure upgrading and the industrial structure rationalization. However, most of the previous research literature only considered industrial structure upgrading, but neglected the indicator of industrial structure rationalization. In view of this, based on the perspective of energy grade calculation, the equivalent electric calculation is used to calculate the conversion coefficients of various energy sources into equivalent electric; therefore, we can accurately obtain the regional energy consumption structure. Then, an empirical study will be used to analyze the effect of industrial structure changing (including industrial structure upgrading and industrial structure rationalization) on energy consumption structure by using the calculation results above. We hope our research can expand the study in this research area.

Study design

Equivalent electric calculation method

As discussed above, the existing statistical calculation of energy consumption ignored the different quality between different kinds of energy. Based on the perspective of energy grade calculation, this article proposed a new statistical calculation method of energy, which is equivalent electric calculation method (EECM). The foundations are as follows: in practical economic productive activities of the region, several main final consumption energies (including coal, natural gas, petroleum, hydropower, nuclear power, solar energy, wind energy, etc.) are all used to produce electricity. Therefore, electricity is used as a foundation to measure other energy grades. The conversion coefficient through which other energies are converted into electricity is defined as this energy’s grade, which is conversion coefficient of equivalent electricity. According to the definition of the equivalent electric calculation above, the grade of electricity is 1, the grade of other energy can be calculated by conversion of the coefficient of equivalent electricity, the calculation model is as follows

E E_{i} = γ_{i} \times E_{i}

(1)

In this model,

i

is other energy sources which include coal, natural gas, petroleum, hydropower, nuclear power, solar energy, wind energy, etc.

E E_{i}

is the corresponding equivalent electricity per unit of energy

i

γ_{i}

is the conversion coefficient of equivalent electricity.

E_{i}

is the total energy. In order to work out the conversion coefficient of equivalent electricity of other energy sources, the following conversion model can be used according to the second law of thermodynamics

γ_{i} = 1 - \frac{T_{i 0}}{T_{i} - T_{i 0}} \cdot \ln \frac{T_{i}}{T_{i 0}}

(2)

In this model,

T_{i}

is the complete combustion temperature of energy

i

T_{i 0}

is the average outdoor temperature in the region 0 where energy

i

is used. In order to work out the conversion coefficient of equivalent electricity of the municipal district heating energy (including hot water and steam), model (2) should be made the following transformation

\begin{array}{l} γ_{w} = 1 - \frac{T_{w 0}}{T_{g} - T_{h}} \cdot \ln \frac{T_{g}}{T_{h}} \\ γ_{h} = 1 - \frac{T_{h 0}}{T_{s}} \end{array}

(3)

In this model,

γ_{w}

is the conversion coefficient of equivalent electricity of the municipal hot water heating system,

T_{g}

is the temperature of the supplied water,

T_{h}

is the temperature of the returned water,

T_{w 0}

is the average outdoor temperature in region 0,

γ_{h}

is the conversion coefficient of equivalent electricity of the municipal steam heating system,

T_{h 0}

is the average outdoor temperature in region 0,

T_{s}

is the corresponding saturation temperature of steam pressure. All the temperature values above are measured in the thermodynamic units (K).

Index calculation of industrial structure changing

Industrial structure changing includes two aspects:the industrial structure upgrading and the industrial structure rationalization. When measuring the index of industrial structure upgrading, this article references the result of Jiao (2015) and use the ratio of the added value of the tertiary industry to the added value of the secondary industry for representation. The larger the ratio is, the better the industrial structure upgrading is. When measuring the index of industrial structure rationalization, this article references the method of Dono and Thompson (1994) and use the Theil index to represent. The model of Theil index is as follows

T L_{i t} \sum_{j = 1}^{n} [(\frac{Y_{i t j}}{Y_{i t}}) \cdot \ln (\frac{Y_{i t j}}{L_{i t j}} / \frac{Y_{i t}}{L_{i t}})]

(4)

In this model,

i, t, j

are the provinces, time periods, and industries, respectively.

Y_{i t j}

is the added value of industry

j

in year

t

in province

i

Y_{i t}

is the gross regional product in year

t

in province

i

L_{i t j}

is the number of the employed persons in industry

j

in year

t

in province

i

L_{i t}

is the total employed persons in year

t

in province

i

. This index is converse. The lower the Theil index is, the better the industrial structure rationalization is.

Econometric model and statistical interpretation of data

This article mainly studies the influence of industrial structure changing on energy structure consumption, so the econometric model is as follows

\begin{matrix} L N (e n e r g y_{i t}) = c + α_{1} L N (u p g r a d e_{i t}) + α_{2} L N (r a t i o n a l i z a t i o n_{i t}) \\ + β_{1} L N (t e c h n o \log y_{i t}) + β_{2} L N (p r i c e_{i t}) \\ + β_{3} L N (G D P_P_{i t}) + ε_{i t} \end{matrix}

(5)

In this model, $i, t$ are the provinces and time periods, respectively. As the dependent variable in this article, energy consumption structure uses the energy structure cleaning ( $e n e r g y_{i t}$ ) as its index, it is formulated by the ratio of non-coal energy consumption to the total energy consumption, which is different from the existing research (Chen et al., 2011; Wang et al., 2011). The core explanatory variable contains two indexes: industrial structure upgrading ( $u p g r a d e_{i t}$ ) and industrial structure rationalization ( $r a t i o n a l i z a t i o n_{i t}$ ). $c$ is model’s constant term; $ε_{i t}$ is the random disturbance term. Existing research shows that apart from the industrial structure, energy consumption structure had many other influence factors which include technological progress, changes of energy prices, economic development levels and so on (Fan et al., 2012; Liu and Li, 2011; Liu et al., 2016). Therefore, this article chooses the technological progress ( $technolog y_{i t}$ ), energy price ( $p r i c e_{i t}$ ), and economic level ( $G D P_P_{i t}$ ) as the control variables of the regression model. The reasons are as follows: Firstly, the indicators of technological progress use the granted quantity of inventive patent applications as its measurement indicator. Patents include invention patents, utility-model patents and design patents. In this study, we think that only the invention patents directly measure the level of regional technological innovation, and hence the granted quantity of invention patents can reflect the level of regional technological progress. Secondly, the energy-price index is measured by the purchasing fuel and power price index. As an important productive factor of regional economic production activities, the price of energy can influence the cost of economic production activity directly, and then influences the choice of economic production activities for the type of energy consumption, that is, the choice of energy consumption structure. Thirdly, the economic indicators are measured by using per capita GDP. Based on the existing literature, we can see that research about the effect of economic development level on energy consumption structure reducing to the level in 2003 do not have unanimous conclusions. On the one hand, existing research results show that higher levels of economic development can promote the exploration and utilization of clean energy such as wind power, solar energy and so on, thus promoting the cleaning of energy consumption structure (Zeng et al., 2009). On the other hand, the excessive growth of economic level will also lead to spring back problems of energy consumption (Shao et al., 2013; Zha and Zhou, 2010). When economic growth is greater than the growth of clean energy supply, economic growth will not be conducive to the cleaning of energy consumption structure.

The empirical results and analysis

Calculating results of conversion coefficient of equivalent electricity of partial energy

According to the statistical results of Chinese regional final energy consumption of physical quantity from the regional energy balance table in “China Energy Statistical Yearbook”, combined with the EECM, this study chooses the main 20 types of final energy consumption to measure its conversion coefficient of equivalent electricity. Taking Beijing in 2013 as an example, we choose the annual average outdoor temperature 12.36°C as the outdoor reference temperature. The calculating results of conversion coefficient of equivalent electricity of the main 20 types of final energy consumption are shown in Table 1.

Table 1.

The conversion coefficient of equivalent electricity of the 20 types of final energy consumption.

Energy types	Final consumption energy	Conversion coefficient of equivalent standard coal		Energy grade	Conversion coefficient of equivalent electricity
Energy types	Final consumption energy	Coal equivalent calculation method	Calorific value calculation method	Energy grade	Conversion coefficient of equivalent electricity
Coal types	Raw coal	0.7143 kgce/kg	0.7143 kgce/kg	0.4263	2.4785 kWh/kg
	Cleaned coal	0.9000 kgce/kg	0.9000 kgce/kg	0.4263	3.1229 kWh/kg
	Other washed coal	0.2850 kgce/kg	0.2850kgce/kg	0.4263	0.9915 kWh/kg
	Briquette	0.6000kgce/kg	0.6000kgce/kg	0.4263	2.4785 kWh/kg
Coke types	Coke	0.9714 kgce/kg	0.9714 kgce/kg	0.4803	3.7977 kWh/kg
	Coke oven gas	0.5714kgce/m³	0.5714kgce/m³	0.4803	2.2338 kWh/m³
	Other coal gas	0.3570 kgce/m³	0.3570 kgce/m³	0.4803	2.2338 kWh/m³
	Other coking products	1.3000 kgce/kg	1.3000 kgce/kg	0.4803	3.7977 kWh/kg
Petroleum types	Crude oil	1.4286 kgce/kg	1.4286 kgce/kg	0.6417	7.4617 kWh/kg
	Gasoline	1.4714 kgce/kg	1.4714 kgce/kg	0.6417	7.6855 kWh/kg
	Kerosene	1.4714 kgce/kg	1.4714 kgce/kg	0.6417	7.6855 kWh/kg
	Diesel	1.4571 kgce/kg	1.4571 kgce/kg	0.6417	7.6109 kWh/kg
	Fuel oil	1.4286 kgce/kg	1.4286 kgce/kg	0.6417	7.4617 kWh/kg
	Liquefied petroleum gas	1.7143 kgce/kg	1.7143 kgce/kg	0.6417	8.4604 kWh/kg
	Refinery off-gas	1.5714 kgce/kg	1.5714 kgce/kg	0.6417	8.2079 kWh/kg
	Other petroleum products	1.2000 kgce/kg	1.2000 kgce/kg	0.6417	8.4604 kWh/kg
Natural gas types	Natural gas	1.2143 kgce/m³	1.2143 kgce/m³	0.6417	6.9464 kWh/m³
Heating types	Hot water	0.0341 kgce/MJ	0.0341 kgce/MJ	0.1688	0.1688 kWh/MJ
Heating types	Steam	0.0341 kgce/MJ	0.0341 kgce/MJ	0.2496	0.2496 kWh/MJ
Electrical types	Electricity	0.3619 kgce/kWh	0.1229 kgce/kWh	1.0000	1.0000 kWh/kWh

From Table 1, on the basis of EECM, we can see that the highest grade energy in the final energy consumption in Beijing in 2013 are electrical energies with a grade of 1.0000; the second is natural gas energies and petroleum energies, both reached the grade of 0.6417; the following are coke energies and coal energies, their grades are 0.1803 and 0.4263, respectively; the lowest grade energy is the thermal energy, the grades of municipal hot water heating and municipal steam heating are only 0.1688 and 0.2496, respectively. Comparing the measured results in the third, fourth, and fifth columns in Table 1, it can be seen that there is a big difference in calculation between the EECM and traditional energy statistical method. The EECM in this study has the following advantages: First, both the coal equivalent calculation method and calorific value calculation method ignore the difference between different types of energies, and it is easy to overestimate low-grade energy in the process of energy statistical calculation. But the EECM can unite two essential factors of “quality” and “quantity” of all kinds of energy, and hence the calculation is more exact. Second, for new types of energy such as solar, wind and so on, using EECM can not only avoid the complex accounting process of converting it into standard coal but also conform to the actual utilization process of new energies because the main utilization of these new energies is generated electricity.

Descriptive statistics

This study focuses on the effect of the regional industrial structure changing on the regional energy consumption structure. We take the regional energy consumption structure as the research object and choose panel data which are from 30 mainland Chinese province, autonomous regions, and municipalities between 2003 and 2013 as samples (due to the lack of statistical data on total energy consumption and fuel/power purchase price index over the years in Tibet, we only selected statistical data in 30 provinces as samples), there are about 330 group observations. The raw sample data are from “China Statistical Yearbook” between 2004 and 2015, “China Energy Statistical Yearbook” between 2004 and 2015, and “China’s Labor Statistical Yearbook” between 2004 and 2015, per capita GDP use this regional consumer price index in former years reducing to the level in 2003 as the standard values of base period.

The calculation results of the maximum/minimum/mean/standard deviation/skewness/kurtosis of each statistical indicator are shown in Table 2. It can be seen from this table that the absolute values of the skewness of each statistical indicator are less than 3. The absolute values of the kurtosis are all less than 10, which generally meet the criteria of normal distribution and can be used for subsequent data analysis.

Table 2.

Descriptive Statistics.

Variables	No.	Minimum	Maximum	Mean	Std. deviation	Skewness		Kurtosis
Variables	Statistic	Statistic	Statistic	Statistic	Statistic	Statistic	Std. error	Statistic	Std. error
Coal Consumption (10⁴ tons)	330	326	40,233	11,115.138	8178.384	1.321	0.134	1.232	0.268
Total energy consumption (10⁴ tce)	330	742	38,899	11,115.282	7488.355	1.174	0.134	1.087	0.268
Value added of the primary industry/(100 million yuan)	330	46.15	4742.63	1151.501	937.1	1.109	0.134	0.934	0.268
Added value of the secondary industry/(100 million yuan)	330	151.16	29,427.49	5822.202	3773.404	1.879	0.134	2.766	0.268
Added value of the Tertiary Industry/100 million yuan）	330	137.84	29,688.97	4711.372	3819.165	2.28	0.134	2.34	0.268
Gross Regional Product/100 million yuan）	330	385.34	62,163.97	11,685.075	9116.006	1.999	0.134	2.721	0.268
Employed persons of the primary industry/(10⁴ persons)	330	36.34	3331.86	988.262	623.418	0.79	0.134	0.159	0.268
Employed persons of the secondary industry/(10⁴ persons)	330	34.8	2563.5	644.234	571.524	1.257	0.134	0.765	0.268
Employed persons of the tertiary industry/(10⁴ persons)	330	76.4	2275.63	795.21	488.833	0.652	0.134	−0.187	0.268
Total employed persons/(10⁴ persons)	330	254.26	6387	2427.711	1201.224	0.635	0.134	−0.53	0.268
Domestic inventive patent applications (piece)	330	17	22,153	1967.73	3522.173	2.411	0.134	2.861	0.268
The purchasing price for fuel and power	330	85.3	123.9	105.625	6.58	−0.343	0.134	−0.24	0.268
Per capita gross regional product (yuan)	330	753.54	100,105	27,464.334	12,122.144	1.351	0.134	1.699	0.268

The regression analysis

In order to avoid spurious regression, this study firstly uses the Levin, Lin & Chut test, the Fisher-ADF test, and the Im, Pesaran and Shin test to test the unit root of each variable. The results show that the logarithmic index of energy structure cleaning is a zero-order unit integer, it is lower than the order of unit integer of all explanatory variables such as industrial structure upgrading, industrial structure rationalization and so on (due to space limitation, we will not present all the orders). This means that the sample is suitable for regression analysis.

Preliminary regression analysis

In Table 2, the explained variable is the index of regional energy structure cleaning, and the five indicators are: industrial structure upgrading, industrial structure rationalization, technological progress, energy price, and economic level as the explaining variable. The results of regression analysis are shown in Table 3.

Table 3.

An effect of industrial structure changing on energy consumption structure.

Variable	(I)		(II)		(III)		(IV)
Variable	Fixed effect	Random effect	Fixed effect	Random effect	Fixed effect	Random effect	Fixed effect	Random effect
_c	−0.2324	−0.2174	−0.3247	−0.2995	−0.3146	−0.3007	−0.3270	−0.3034
upgrade_it	0.0940**(2.0395)	0.1058**(1.9627)	0.0754**(2.3708)	0.0861**(2.1208)	0.0762**(2.2894)	0.0853**(2.1924)	–	–
rationalization_it	0.1706*(1.7206)	0.1627*(1.7007)	0.1429** (2.2114)	0.1209**(2.0812)	–	–	0.1459*(1.7908)	0.1223*(1.6806)
technology_it	–	–	0.1507**(1.9827)	0.1724**(2.0877)	0.1527***(3.2114)	0.1738***(2.8974)	0.1555**(2.3562)	0.1783**(2.1709)
price_it	–	–	−0.2438*(1.7487)	−0.2208*(1.6329)	−0.2471*(1.7740)	−0.2258*(1.7067)	−0.2529*(1.7412)	−0.2297**(1.6905)
GDP_P_it	–	–	0.0791 (1.4061)	0.0615 (1.2109)	0.0798 (0.8043)	0.0604 (0.7426)	0.0791 (1.1524)	0.0610 (1.0825)
F	27.3046***	24.0917***	34.2534***	25.1906***	41.2307***	34.0069***	35.0168***	29.8729***
Adj-R²	0.2342	0.2094	0.3927	0.3211	0.3752	0.3364	0.3524	0.3019
Hausman test	3.7324(0.4218)		2.3526(0.2147)		4.0674(0.1907)		3.8376(0.2183)

Note: Numbers in parentheses in the last line indicate adjoint probability, numbers in other parentheses indicate T-statistic of relevant variable; ***, **,* mean passing 1%, 5%, and 10% significance testing. Similarly hereinafter.

From the regression analysis, we can see that in models (I)–(IV), the regression coefficient of industrial structure upgrading and industrial structure rationalization shows a strong stability. This means that the regression results of this model are dependable strongly. The adjoint probability of Hausman’s test statistics in models (I)–(IV) is 0.4218, 0.2147, 0.1907, and 0.2183, respectively. Therefore, the regression results of random effect model are more suitable for this sample. The following analyses are based on the regression results of random effect model. Firstly, the industrial structure changing has a significant positive effect on the regional energy structure cleaning. The indicators of industrial structure upgrading and industrial structure rationalization all make a significant positive effect on energy structure cleaning. The model (II) shows that the regression coefficient of industrial structure upgrading is 0.0861 with a significance under 5% (sig). This means, when the regional industrial structure upgrading increases 1%, the regional energy structure cleaning will increase 0.0861%. The regression coefficient of industrial structure rationalization is 0.1209 with a significance under 5% (sig). This means, when the regional industrial structure rationalization increases 1%, the regional energy structure cleaning will increase 0.1209%. Meanwhile, we find that the influence coefficient of industrial structure rationalization is higher than that of industrial structure upgrading, this means the rational development of industrial structure and the optimized construction of industry chain play a greater role in the adjustment of the regional energy structure. Secondly, technological progress has a significant positive effect on regional energy structure cleaning. When the regional technological progress increases 1%, the regional energy structure cleaning will increase 0.1724%; Energy-price changing has a negative effect on the regional energy structure cleaning, its regression coefficient is −0.2208, and it passed a significant 10% test. The decline in energy price is conducive to the improvement of the energy structure cleaning. During the sample period, the level of economic development has no significant effect on the index of energy structure cleaning.

Robust regression analysis

In order to analyze the effect of industrial structure changing on the robustness of energy consumption structure, this study chose the proportion of generated energy of clean energies such as Chinese regional hydropower, nuclear power, wind power and so on in the total amount of regional electricity production as the index of energy structure cleaning to do regression analysis. The results are shown in Table 4.

Table 4.

Robust analysis of industrial structure changing affecting energy consumption structure.

Variable	(V)	(VI)	(VII)	(VIII)
Variable	Random effect	Random effect	Random effect	Random effect
$c$	−0.3207	−0.3712	−0.3419	−0.3547
$u p g r a d e_{i t}$	0.0523**(2.1058)	0.0434**(2.2287)	0.0411**(2.3527)	–
$r a t i o n a l i z a t i o n_{i t}$	0.0804*(1.7274)	0.0609**(1.9049)	–	0.0603*(1.7098)
technology_it	–	0.0873**(2.5130)	0.0840***(3.0923)	0.0883***(3.5718)
$p r i c e_{i t}$	–	−0.1116*(−1.7130)	−0.1090*(−1.7876)	−0.1137*(−1.6498)
GDP_P_it	–	0.0309 (0.6145)	0.0290 (0.6192)	0.0300 (0.7814)
F	32.8188***	34.3158***	46.3257***	40.6942***
Adj-R²	0.2317	0.3024	0.3648	0.3936

Note: The results of Hausman test indicate that random effects model is more suited to this sample. Due to space limitation, this study only lists the regression results of random effects model.

The results show the following: Firstly, the indicators of industrial construction upgrade, industrial construction rationalization and technological progress all have a significant positive effect on the generated energy ratio of clean energy. In model VI, the regression coefficients are 0.0434 and 0.0609, respectively, which mean when industrial structure upgrading, industrial structure rationalization and technological progress increase 1%, the generated energy ratio of clean energy will increase 0.0434 and 0.0609, respectively. Secondly, the regression coefficients of energy-price changing indicator are all negative and they passed a 10% significance test, which means the decline in energy prices during the sample period is conducive to increasing the proportion of clean energy use. Thirdly, the economic level indicator has no significant effect on the generated energy ratio of clean energy. Therefore, the significance test results and the sign of the regression coefficient in models (V)–(VIII) are consistent with the models (I)–(IV), which means the above empirical analysis results have good robustness.

Conclusion and enlightenment

Based on the perspective of energy grade, this paper firstly gave the EECM of energy statistical calculation, integrates the “quantity” and “quality” of various energies into energy statistical calculation. Then, by using the results of the energy consumption calculation above, the energy consumption structure is calculated. By combining with the data of actual economic production activities in 30 Chinese provinces between 2003 and 2013, the effect of the regional industrial structure changing on the energy consumption structure is empirically analyzed. Through the analysis above, the main conclusions are as follows:

Firstly, there is a significant difference in five different energy grades in final energy consumption in China, the five energies include coal, petroleum, natural gas, heating power, and electricity, and their energy grades from high to low is electricity, natural gas, petroleum, coal, heating power, respectively. Both the traditional coal equivalent calculation method and the calorific value calculation method ignore the difference of grade between different energies, and both the methods overestimate low-grade energies in the process of energy statistical calculation.

Secondly, both the industrial structure upgrading and the industrial structure rationalization have significant and steady positive effect on the regional energy structure cleaning, and the promotion of the regional industrial structure rationalization to the regional energy structure cleaning is greater, when its index increases 1%, the regional energy structure cleaning will increase 0.1209%. The industrial structure upgrading and the industrial structure rationalization can also promote the increase of the consumption proportion of the regional new energy (including hydropower, nuclear power, wind power, solar electrical energy generation and so on), when the index of the two indicators increase 1%, the regional consumption proportion of the new energy will increase 0.0434% and 0.0609%, respectively.

Thirdly, in controlling variables, the technological progress indicator has significant and steady positive effect on the regional energy structure cleaning and the consumption proportion of the regional new energy, but energy-price changing has a negative effect on the two measuring indicators of energy consumption structure, the decrease of energy price is conducive to the increase of the energy structure cleaning and the consumption proportion of new energy. Finally, during the sample period, the developing level of regional economy has no significant effect on energy consumption structure.

The enlightenment of this paper is as follows: the energy grades represent the “quality” of energy, it can reflect the function of different types of energies. Thus, we suggest that governments establish and practice a set of “equivalent electricity” methods to calculate the production, transportation, and consumption of all kinds of energies, and calculate all kinds of energy grades, so as to better connect energy suppliers and energy demanders, and can effectively avoid unreasonable use of “high-grade energy and low-grade energy use”. The significant positive effect of industrial structure rationalization on the development of the regional energy structure cleaning is higher than that of the industrial structure upgrading. Therefore, under the economic environment in which Chinese industrial structure upgrading is becoming more accelerated, the adjustment of the regional industrial structure should not blindly pursue industrial structure upgrading, it should pay more attention to the development of industrial structure rationalization and the supporting construction in the upstream and downstream industries in the industrial chain in order to achieve the sustainability of industrial structure upgrading. In addition, governments should increase investment in technological innovation and issue various preferential policies to encourage enterprises to develop green manufacturing technologies. At the same time, the government should actively adopt steps such as energy-price subsidies and other measures to reduce enterprise’s energy consumption price. Moreover, in the case of existing energy-price subsidies, the price subsidy policy should be biased toward the high-grade and cleaning energy prices to promote the development of energy structure cleaning.

Footnotes

Acknowledgments

This research is conducted through the Technological Economy and Innovation Management Institute in the College of Economics and Management, Nanjing University of Aeronautics and Astronautics.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by China Scholarship Council, the Funding of Jiangsu Innovation Program for Graduate Education and the Fundamental Research Funds for the Central Universities.

References

Chen

Liu

Shi

et al . (2011) The characteristics of CO₂ emission based on the structure of energy consumption – A case study in Ningbo, China. Research Journal of Environmental & Earth Sciences 3(14): 2928–2935 (in Chinese).

Dono

Thompson

(1994) Explaining changes in Italian consumption of meat: Parametric and non-parametric analysis. European Review of Agricultural Economics 21(2): 175–198.

Duarte

Restuccia

(2010) The role of the structural transformation in aggregate productivity. The Quarterly Journal of Economics 125(1): 129–173.

Fan

Wang

Zhang

(2012) Analysis of the influence factors of the primary energy consumption structure under the target of low-carbon economy. Resources Science 34(4): 696–703.

Gan

Zheng

(2009) An empirical study on change of industrial structure and productivity growth since the reform and opening-up – A test for the structure-bonus hypotheses from 1978 to 2007 in China. China Industrial Economics 2: 55–65.

Gan

Zheng

(2011) An empirical study on the effects of industrial structure on economic growth and fluctuations in China. Economic Research Journal 5: 4–16.

Guo

Sun

Dai

(2016) Effect of population structure change on carbon emission in China. Sustainability 8(3): 225.

Jang

Liu

Yang

(2006) A study on calculation method of different energy source in energy statistics. Energy of China 28(6): 5–8 (in Chinese).

Jiao

(2015) Does geographic agglomeration of production factors influence industrial structure change. Statistical Research 32(8): 54–61 (in Chinese).

10.

(2012) Ecological total-factor energy efficiency of regions in China. Energy Policy 46: 216–224.

11.

Liu

Chen

Wang

et al . (2016) Evolution of China’s urban energy consumption structure – A case study in Beijing. Energy Procedia 88: 88–93.

12.

Liu

Dong

(2011) Spatial analysis on the contribution of industrial structure change to regional energy efficiency: A case study of 31 provinces in mainland China. Journal of Natural Resources 26(12): 1999–2011.

13.

Liu

(2011) Improving energy consumption structure: A comprehensive assessment of fossil energy subsidies reform in China. Energy Policy 39(7): 4134–4143.

14.

Peng

(2013) Study on the dynamic effects of industrial structure on economic fluctuation. Industrial Economics Research 3: 94–100.

15.

Shao

Yang

Huang

(2013) Theoretical model and experience from China of energy rebound effect. Economic Research Journal 2: 96–109.

16.

Song

Zhou

(2015) Analysis of carbon emissions and their influence factors based on data from Anhui of China. Computational Economics 46(3): 359–374.

17.

Sun

Zhou

et al . (2015) The structural adjustment and energy related carbon emission: Review on the target of emission reduction achieved. Journal of Applied Statistics and Management 34(1): 59–74 (in Chinese).

18.

Wang

(2011) Analyze the relationship between energy consumption and economic growth in China. Energy Procedia 5: 974–979.

19.

Zeng

Zheng

Rao

(2009) Analysis of the correlative relationships between energy consumption structure and economic growth in China. Soft Science 23(8): 65–68.

20.

Zha

Zhou

(2010) The research on China’s energy efficiency rebound effect based on CGE model. Journal of Quantitative Techniques and Economics 12: 39–53 (in Chinese).

21.

Zhang

(2015) Analysis on influence factors of China’s provincial industrial structure upgrading from a spatial econometrics perspective. Statistical Research 32(10): 32–37.

22.

Zhang

Choi

(2013) Total-factor carbon emission performance of fossil fuel power plants in China: A metafrontier non-radial Malmquist index analysis. Energy Economics 40(2): 549–559.

23.

Zhang

Zhou

Choi

(2013) Energy efficiency, CO₂ emission performance and technology gaps in fossil fuel electricity generation in Korea: A meta-frontier non-radial directional distance function analysis. Energy Policy 56(2): 653–662.

24.

Zhou

Zhang

Zhou

et al . (2013a) Modeling economic performance of interprovincial CO₂ emission reduction quota trading in China. Applied Energy 112(16): 1518–1528.

25.

Zhou

Xing

Fang

et al . (2013b) Environmental efficiency analysis of power industry in China based on an entropy SBM model. Energy Policy 57(10): 68–75.