Analysis on Influencing Factors Decomposition and Decoupling Effect of Power Carbon Emissions in Yangtze River Economic Belt

Abstract

The power industry is one of the main industries of energy consumption and carbon emissions in the Yangtze River Economic Belt, in order to accelerate the development of green and low-carbon power. This paper takes the power industry in the Yangtze River Economic Belt and the upper, middle and lower reaches from 2000 to 2020 as the research object. Based on the four absolute factors of economy, energy consumption, population and output, the generalized divisia index method (GDIM) is constructed. Combining the decoupling model with the GDIM method, a decoupling effort model is constructed based on the DPSIR framework to incorporate electricity output and economic scale into economic drivers. The main findings are as follows: The lower reaches provide the main contribution area of power carbon emissions, and the high value of carbon emissions tends to migrate eastward. Economy, output and energy consumption scale are the main factors leading to the increase in carbon emissions in the Yangtze River Economic Belt, upper, middle and lower reaches, while carbon intensity and output carbon intensity are the key factors curbing carbon emissions. Among them, economic scale is the largest positive driving factor of carbon emissions in the middle and lower reaches, and output carbon intensity is the largest negative driving factor in the upper reaches. The total decoupling effect index in the upper reaches of the Yangtze River Economic Belt increased the most and achieved decoupling of carbon emissions between 2014 and 2020. From 2016 to 2020, the total decoupling effect index of the upper, middle and lower reaches of the Yangtze River Economic Belt fluctuates within a certain range, and the total decoupling effect index of the power industry has entered a certain pressure period.

1. Introduction

The 2021 IPCC Sixth Assessment Report states that the global carbon dioxide concentration in 2019 is higher than at any time in at least 2 million years [1]. Increased carbon emissions cause global warming, which poses a certain threat to the normal survival and development of human beings [2]. In 2020, China completed its commitment to the international community to reduce carbon emission intensity on the eve of the 2009 Copenhagen Climate Change Conference [3]. In 2020, China announced that it will strive to achieve carbon peak and carbon neutrality in 2030 and 2060 [4], and China has entered a critical period of achieving the strategic goal of carbon peak and carbon neutrality. In 2020, China’s 14th Five-Year Plan pointed out that China should speed up its move towards green and low-carbon development. As a result, China’s carbon peak, carbon neutrality target and green low-carbon development plan will impose higher requirements on China’s power industry to control and reduce carbon emissions in the future. China has abundant reserves of coal resources and long-term fossil energy coal-based power generation; carbon emissions caused by China’s electricity production accounted for 42.52% of the country’s total carbon emissions in 2020. In 2021 China’s electricity demand grew by 10%, higher than the economic growth rate of 8.4% in the same period, of which 56% of the new electricity demand was filled by coal [5]. With the steady development of economic recovery in the post-epidemic era, the pressure between China’s economic development and power energy conservation and emission reduction still exists.

As one of the three strategic regions in China [6], the Yangtze River Economic Belt has strong development potential relying on the golden waterway. In 2019, the carbon emissions from electricity in the Yangtze River Economic Belt accounted for 31.13% and 35.32% of China’s total carbon emissions from electricity and the Yangtze River Economic Belt, respectively. The 14th Five-Year Plan points out the construction of a clean, low-carbon, safe and efficient energy system. The power industry in the Yangtze River Economic Belt is further moving towards green and sustainable development, which promotes the construction of clean and low-carbon energy in China. In addition, the 14th Five-Year Plan for the Development of the Yangtze River Economic Belt puts ecological environmental protection in the first place. The country’s environmental protection requirements for the Yangtze River Economic Belt continue to improve, and with the steady increase in economic recovery in the post-epidemic era the power industry is facing further transformation pressure. The further green and sustainable development of the power industry in the Yangtze River Economic Belt has great significance to the demonstration and guidance of China’s ecological civilization construction and high-quality development of the power industry.

The main research contents and purposes are as follows: Firstly, the research status of power carbon emission factor decomposition and decoupling is systematically summarized. Explore the spatial and temporal variation characteristics of power carbon emissions in the Yangtze River Economic Belt and the differences between power carbon emissions in various regions. Secondly, the GDIM model of power carbon emissions is constructed. The influencing factors of power carbon emissions are decomposed to clarify the influence and mode of action of influencing factors on the change in power carbon emissions. Then, the power carbon emissions decoupling effort model is constructed to explore the change characteristics and differences in the total decoupling effect index of power in the Yangtze River economic belt, and the mechanism of decoupling effect of power carbon emission influencing factors on carbon emission decoupling. Finally, the contribution of influencing factors and the change in decoupling effect in different regions of the Yangtze River Economic Belt are compared over different periods. This is combined with the actual search for the power industry effective path of green sustainable development.

The contributions are as follows: Firstly, this paper focuses on the Yangtze River Economic Belt, a strategic development area in China, and studies the carbon emissions of the upper, middle and lower reaches power simultaneously. Through the comparative analysis of power carbon emissions, influencing factors decomposition and decoupling effects in different regions of the Yangtze River Economic Belt, the research on power carbon emissions in the Yangtze River Economic Belt is further supplemented. Secondly, the GDIM method considers the interdependence between the influencing factors of power carbon emissions, taking into account the influence of absolute factors and relative factors on the carbon emissions of the power industry. This paper further uses four absolute factors and six relative factors to explore the contribution to power carbon emissions. The influencing factors of power carbon emissions are considered more comprehensively and the result analysis is more detailed. Finally, the decoupling effort method can objectively measure the actual decoupling effect of carbon emissions from electricity caused by the carbon emission reduction efforts of the government and the power industry. The current research usually uses only one economic driving factor to construct the decoupling effort model. In this paper, electricity output and economic scale are included as economic factors in the decoupling effort model of carbon emissions, and then the decoupling state between economic benefits and environmental pressure is explored, so that the decoupling effect is more convincing. In addition, this paper uses 20 kinds of electricity energy consumption to calculate carbon emissions, basically covering all kinds of power energy consumption. GDP is calculated at 2000 constant prices in 11 provincial administrative regions, which is a more objective measure of how the economy affects carbon emissions from electricity. Based on the scholar Vaninsky, the GDIM model of power carbon emissions is introduced in detail. At the same time, a detailed description is provided of the power carbon emissions decoupling effort model construction principle.

2. Literature Review

Structural decomposition analysis (SDA) and Index decomposition analysis (IDA) are commonly used to decompose factors. SDA is based on input–output tables for factor decomposition, and the IDA method requires the use of sectoral summation data for factor decomposition [7]. The structural decomposition method decomposes the dependent variable into the sum of independent variables by an input–output model, but the structural decomposition method has high requirements for data. As early as 1970, Leontief [8] applied the input–output structural decomposition method to the field of energy and environment, and then Ma et al. [9], Luo et al. [10], and Jiang et al. [11] used the structural decomposition method to investigate China Power Carbon emissions, the hidden carbon emissions of China’s power sector, carbon emissions of China’s power and heating industries. The common decomposition models used in the Index decomposition method are the Divisia decomposition model and the Laspeyres decomposition model. As early as 1996, Shrestha and Timilsina [12] used the Divisia method to study the influencing factors of carbon dioxide intensity changes in the power sector in related regions of Asia. Sun et al. [13] used Laspeyres decomposition to study the regional differences in the factors influencing power carbon emissions between different provinces in China. As the research progressed, Ang et al. [14] proposed the Logarithmic Mean Divisia Index (LMDI), which does not produce residuals after factor decomposition, while Yang and Lin [15], Mai et al. [16], and Li et al. [17], Zhao et al. [18], and Wu et al. [19] selected population factors as absolute quantity factors, and Cao and Jiang [20], He et al. [21] and Li et al. [22] chose economic factors as absolute quantity constructs to explore the factors influencing power carbon emissions using LMDI model. Vaninsky [23,24] proposed GDIM method, which can simultaneously quantify the impact of multiple absolute and relative factors on the target variable, and the GDIM has been popularized and used in industries such as industry, manufacturing, and transportation, as shown in

Table 1. Summary of the research on the generalized divisia index method.

Author (Year)—Research Industry	Absolute Variable	Main Promoting Factors	Main Inhibiting Factors
Shao et al. [28] (2016)—mining	output scale, energy	output scale	carbon intensity
Ma et al. [29] (2019)—industrial	output scale, technological progress, energy consumption, population	output scale, technological progress	output carbon intensity, technological progress carbon intensity
Wen et al. [30] (2022)—industrial sub-sectors	energy consumption; economic output, investment scale	investment scale	carbon intensity of investment, energy intensity
Wang et al. [31] (2021)—Renewable energy	gross domestic product, total energy consumption, renewable energy	renewable energy scale	carbon intensity of renewable energy
Zhang et al. [32] (2020)—Energy	energy consumption, GDP, population, fixed asset investment	investment sale expansion	carbon intensity of investment
Wang et al. [33] (2018)—Transportation	transportation added value, energy consumption, population size	added value of transportation	energy carbon emission intensity
Wang et al. [34] (2021)—information and communication technology	value added, gross investments, ICT investment	emission intensity of ICT investments	structure and efficiency of ICT investments
Xu and Chen [35] (2022)—nonferrous metal	total energy consumption, industrial value added, investment in fixed assets	output scale	carbon intensity of output, carbon intensity of investment
Jin and Han [36] (2021)—manufacturing	energy consumption, value added, fixed asset investment	fixed asset investment	investment carbon intensity

. In addition, Zhu et al. [25], Yan et al. [26], and Wang et al. [27] constructed a generalized divisal index method to explore the factors influencing carbon emissions in the power industry, as shown in

Table 2. Summary of the research on the generalized divisia index method in power industry.

Author (Year)	Region	Absolute Variable	Result
Zhu et al. [25] (2018)	China’s power industry	added value, energy consumption, population	The output scale(GDP) is the main factor leading to the increase in carbon emissions.
Yan et al. [26] (2019)	Beijing–Tianjin-Hebei region	electricity output, energy consumption	Electricity demand is the main factor that promotes the increase in CO2 emissions.
Wang et al. [27] (2019)	China eight regional power sectors	GDP, energy consumption, scale of output(power generation).	GDP and output scale are the main factors affecting the carbon emissions of the eight regional power sectors.

. In the research field of GDIM model power carbon emissions, scholars have concluded that power output [26,27] and economic output [25,27] are the most important drivers of power carbon emissions, while power output carbon intensity [26,27] and economic output carbon intensity [25,27] are the key inhibitory effects of power carbon emissions.

OECD decoupling, Tapio decoupling and Decoupling based on complete decomposition techniques are the more commonly used decoupling methods. OECD decoupling is a concept proposed by OECD in 2002 [37], which describes the degree of responsiveness between driving forces and environmental pollution pressures. Tapio decoupling is a concept first introduced by scholar Tapio [38], which allows for a causal chain decomposition of decoupling indicators [39]. Xie et al. [40], and Raza and Lin. [41] used Tapio decoupling to explore the decoupling relationship between power carbon emissions and the economy. Chen et al. [42], Wang et al. [43], and Liu et al. [44] used Tapio decoupling to investigate the decoupling relationship between power carbon emissions and electricity. With the deepening of the research, scholars combine the factor decomposition method with the decoupling model. Zhao et al. [45] and Zha et al. [46], through Tapio decoupling combined with LMDI method, measured the decoupling state of China’s economic development and carbon emissions, and Chengdu’s tourism economic growth and carbon emissions. In addition, Diakoulaki and Mandaraka [47] combined Laspeyres method and decoupling method to construct the decoupling effort method to explore the effect of carbon emission reduction efforts made by EU countries. After that, Xu et al. [48] constructed a decoupling effort model for manufacturing carbon emissions in China based on the combination of the Laspeyres decomposition method and OECD decoupling. With the further research on the factor decomposition method, He and Liu. [49], Yuan et al. [50], and Liu et al. [51] used LMDI combined with OECD proposed DPSIR theory to construct a decoupling effort model of carbon emissions in the Beijing–Tianjin–Hebei logistics industry, Chinese transport industry and Tibetan transportation industry. Some scholars also used the GDIM decomposition method in combination with OECD decoupling to explore the decoupling of carbon emissions in related industries. Ma et al. [29] and Wang et al. [27] used the GDIM method combined with OECD decoupling to construct a decoupling effort model to explore the decoupling effect of carbon emissions in China’s industrial and power sectors.

By combing the literature, it can be seen that there are few sources on power carbon emissions in the Yangtze River Economic Belt of China’s strategic region. The structural decomposition method and logarithmic mean Divisia index decomposition method are widely used in power carbon emissions. Vaninsky (2013–2014) proposed the GDIM model in the field of power carbon emissions, but the number of basic model indicators is limited. The decoupling effort method is constructed based on DPSIR framework by combining decoupling model and index decomposition method. It can objectively measure the actual effect that government and industry have emission reduction. The decoupling effort model is currently used by scholars, usually only for the economy or output into an economic factor, or both if considered at the same time they will be more objective.

3. Methods

3.1. Carbon Emission Calculation of Power Industry

The method proposed by IPCC in 2006 is used to calculate the power carbon emissions. The calculation formula is as follows:

$$\mathrm{C}{\mathrm{O}}_2={\displaystyle \sum }_{i=1}^{20}\times NC{V}_i\times CE{F}_i\times CO{F}_i\times 44/12$$

(1)

In the above equation, CO₂ represents the carbon emissions; i represents the types of power energy consumption (20 kinds of energy consumption categories

Table 3. Power consumption energy varieties.

Categories	Coal Total	Petroleum Products Total	Natural Gas
Energy varieties	Raw Coal, Cleaned Coal, Other Washed Coal, Briquettes, Gangue, Coke, Coke Oven Gas, Blast Furnace Gas, Converter Gas, Other Gas, Other Coking Product	Crude oil, Diesel Oil, Fuel Oil, Petroleum Coke, Liquefied Petroleum Gas, Refinery Gas, Other Petroleum Products	Natural Gas, Liquefied Natural Gas

); E is the energy consumption; NCV is the average low level heat generation of electricity energy types; CEF is the carbon content per unit calorific value; COF is the carbon oxidation rate; 44/12 is the carbon molecular weight ratio of CO₂.

3.2. Decomposition Model of Influencing Factors of Carbon Emission in Power Industry

The GDIM method was used to decompose the influencing factors of carbon emission. The relevant variables and their meanings in the decomposition model are shown in

Table 4. Variables involved in the model and their implications.

Variables	Meaning
C	Carbon dioxide emissions
X1 = G	Economic scale (regional GDP)
X2= C/G	Power carbon intensity (power carbon emissions per unit of GDP)
X3 = E	Power industry energy consumption
X4 = C/E	Energy carbon intensity (Carbon emissions per unit of energy consumption)
X5 = P	Population (Year-end population by region)
X6 = C/P	Population carbon intensity (power carbon emissions per unit population)
X7 = T	Electricity output scale (electricity generation)
X8 = C/T	Output carbon intensity (carbon emissions per unit of electricity generation)
X9 = G/P	per capital gross regional product
X10 = E/G	Power energy intensity (power energy consumption per unit of GDP)

.

In this paper, the ten factor variables decomposed from power carbon emissions (target variables) are regarded as continuous differentiable functions of time t. The time is integrated to obtain the contribution of each factor variable to the target variable. The mapping relationship between factor variables and target variables is as follows:

$$C=f(X)=f(X_1,X_2,\cdots ,X_{10})$$

(2)

$$\Delta C=C_T-C_0={\displaystyle \sum _{i=1}^{10}\Delta C\left[X_i\right]}={\displaystyle \underset{L}{\int }dC}$$

(3)

In Equations (2) and (3), C represents power carbon emissions, L is the time span, and X_i represents the factor variable. The symbol $\Delta C$ represents the change in power carbon emission C from the base time 0 to the current time T, and $\Delta C\left[X_i\right]$ represents the contribution of factor variable $X_i$ to $\Delta C$. The element of diagonal matrix dX is $d{X}_i=X_i^{\prime }dt$.

Then:

$$X_i=X_i(t)$$

(4)

$$f_i^{\prime }=\frac{\partial f(X_1,X_2,\cdots ,X_{10})}{\partial X_i}$$

(5)

$$\Delta C={\displaystyle \underset{L}{\int }{f}_1^{\prime }d{X}_1}+{\displaystyle \underset{L}{\int }{f}_2^{\prime }d{X}_2}+\cdots +{\displaystyle \underset{L}{\int }{f}_{10}^{\prime }d{X}_{10}}$$

(6)

Then

$$\Delta C\left[X_i\right]={\displaystyle \underset{L}{\int }{f}_i^{\prime }d{X}_i}={\displaystyle {\int }_{t_0}^{t_1}{f}_i^{\prime }}{X}_i^{\prime }dt$$

(7)

where $t_0$ and $t_1$ represent the reference time and the current time, respectively.

In vector form, Equation (7) is expressed as:

$$\Delta C={\displaystyle \underset{L}{\int }\nabla C^T\cdot dX}$$

(8)

where $\Delta C$ is a row vector, $\nabla C$ is a gradient vector, and T represents a transpose.

Vaninsky added Equation (9) to the GDIM model to constrain the relationship between the factors of decomposition Equation (11).

$${\Phi }_j\left(X_1,\cdots ,X_{10}\right)=0,j=1,2,\cdots ,5$$

(9)

Equation (9) can be written in vector form:

$$\Phi \left(X\right)=0$$

(10)

In this paper, from the perspective of electricity production, power carbon emissions factors see Equation (11), the model involves specific meaning see Table 4.

$$\begin{array}{c}C=\left(C/G\right)\cdot G=\left(C/E\right)\cdot E=\left(C/P\right)\cdot P=\left(C/T\right)\cdot T\\ G/P=\left(C/P\right)/\left(C/G\right)\\ E/G=\left(C/G\right)/\left(C/E\right)\end{array}$$

(11)

The above equation is simplified as follows:

$$\begin{array}{c}C=X_1{X}_2=X_3{X}_4=X_5{X}_6=X_7{X}_8\\ X_9=X_6/X_2\\ X_{10}=X_2/X_4\end{array}$$

(12)

Further using the GDIM method, the above equation can be transformed into the following form:

$$\left\{\begin{array}{c}C=X_1{X}_2\\ X_1{X}_2-X_3{X}_4=0\\ X_3{X}_4-X_5{X}_6=0\\ X_5{X}_6-X_7{X}_8=0\\ X_1-X_5{X}_9=0\\ X_3-X_1{X}_{10}=0\end{array}\right.$$

(13)

The multiplication of any pair of factors in the first line of Equation (13) does not change the carbon emission of electricity. Lines 2–6 in Equation (13) are the equations that constitute Equation (9). The function C(X) represents the contribution of factor X to the change in carbon emission, and the gradient variable of C(X) represents $\nabla C={\left(C/G,G,0,0,0,0,0,0,0,0\right)}^T$. The Jacobian matrix composed of various influencing factors constructed by Equation (13) can be expressed as follows:

$${\Phi }_X={\left(\begin{array}{cccccccccc}{X}_2& X_1& -X_4& -X_3& 0& 0& 0& 0& 0& 0\\ 0& 0& X_4& X_3& -X_6& -X_5& 0& 0& 0& 0\\ 0& 0& 0& 0& X_6& X_5& -X_8& -X_7& 0& 0\\ 1& 0& 0& 0& -X_9& 0& 0& 0& -X_5& 0\\ -X_{10}& 0& 1& 0& 0& 0& 0& 0& 0& -X_1\end{array}\right)}^T$$

(14)

The change in power carbon emission C can be decomposed into the sum of the contribution of ten influencing factors as follows:

$$\Delta C\left[X\left|\Phi \right.\right]={\displaystyle \underset{L}{\int }\nabla C^T(I-{\Phi }_X{\Phi }_X^{+})dX}$$

(15)

$${\left({\Phi }_X\right)}_{ij}=\frac{\partial {\Phi }_j}{\partial X_i}$$

(16)

In the equation, the vector $\Delta C\left[X\left|\Phi \right.\right]$ represents the factorization of the change in the target index C in the presence of Equation (10). The symbol ${\Phi }_X$ is the Jacobi matrix of $\Phi \left(X\right)$, while ${\Phi }_X^{+}$ represents the generalized inverse matrix of ${\Phi }_X$, ‘+’ represents the generalized inverse matrix, and I represents the unit matrix. If the variables of each influencing factor of ${\Phi }_X$ are linearly independent, then ${\Phi }_X^{+}={\left({\Phi }_X^T{\Phi }_X\right)}^{-1}{\Phi }_X^T$.

According to Equation (15), the change in carbon emissions is decomposed into the sum of the contribution of ten factors to carbon emissions from electricity. Ten variables contribute to $C_{X_1}$, $C_{X_2}$, $C_{X_3}$, $C_{X_4}$, $C_{X_5}$, $C_{X_6}$, $C_{X_7}$, $C_{X_8}$, $C_{X_9}$, $C_{X_{10}}$ variables explained as shown in Table 4. Among them, four are absolute influence factors, $C_{X_1}$, $C_{X_3}$, $C_{X_5}$, $C_{X_7}$ and six relative quantity influence factors, $C_{X_2}$, $C_{X_4}$, $C_{X_6}$, $C_{X_8}$, $C_{X_9}$, $C_{X_{10}}$.

3.3. Power Carbon Emissions Decoupling Effort Model

The decoupling theory is based on the driver-pressure-state-influence-response (DPSIR) framework proposed by OECD in 1993. In this paper, driver refers to the potential causes of environmental damage, GDP and power output as the driving force. Pressure refers to the direct impact of power carbon dioxide emissions on the environment. State refers to the state of the environment under the action of power carbon dioxide emissions. Influence refers to the influence of power carbon dioxide on human and social development. Response refers to the efforts or positive policies made by the government and industry to restrict power carbon emissions to achieve sustainable development.

The change in power carbon emission can be decomposed into the contribution of 10 influencing factors, as shown in Equation (17).

$$\Delta C=\Delta C_{X_1}+\Delta C_{X_2}+\Delta C_{X_3}+\Delta C_{X_4}+\Delta C_{X_5}+\Delta C_{X_6}+\Delta C_{X_8}+\Delta C_{X_9}+\Delta C_{X_{10}}$$

(17)

The power industry X₁ (GDP) and X₇ (Electricity output) are economic driving factors $\Delta Q$, as shown in Equation (18). In this paper government and industry energy efficiency efforts are policies or measures taken to directly or indirectly reduce carbon emissions, such as improving the level of power production, energy efficiency and changing the power structure to reduce power carbon emissions. Therefore, according to the GDIM decomposition results, the emission reduction efforts ($\Delta R$) reflecting the factor government and industry can be indirectly expressed as Equation (19):

$$\Delta Q=\Delta C_{X_1}+\Delta C_{X_7}$$

(18)

$$\Delta R=\Delta C-\Delta Q=\Delta C-\Delta C_{X_1}-\Delta C_{X_7}=\Delta C_{X_2}+\Delta C_{X_3}+\Delta C_{X_4}+\Delta C_{X_5}+\Delta C_{X_6}+\Delta C_{X_8}+\Delta C_{X_9}+\Delta C_{X_{10}}$$

(19)

Where $\Delta R$ represents the policies or measures taken by human beings to limit carbon emissions, which can be understood as ‘response’ factors, $\Delta C_{X_1}$ and $\Delta C_{X_7}$ represent the economic driving factors of carbon emissions, and $\Delta C$ represents the direct pressure factor of the environment. When $\Delta Q>0$, the total decoupling effect D_t of carbon emissions in the power industry can be expressed as Equation (20). When $\Delta Q<0$, the total decoupling effect D_t can be expressed in Equation (21):

$$D_t=-\frac{\Delta R}{\Delta Q}=D_{X_2}+D_{X_3}+D_{X_4}+D_{X_5}+D_{X_6}+D_{X_8}+D_{X_9}+D_{X_{10}}$$

(20)

$$D_t=\frac{\Delta R}{\Delta Q}=D_{X_2}+D_{X_3}+D_{X_4}+D_{X_5}+D_{X_6}+D_{X_8}+D_{X_9}+D_{X_{10}}$$

(21)

In the actual research, the economic factor ($\Delta Q$) is the main factor causing carbon emissions and ($\Delta Q$) is usually driven positively. In this paper, the economic driving factors of ($\Delta Q$) always have a positive impact on carbon emissions caused by electricity. Scholars Diakoulaki and Mandaraka proposed a decoupling effort formula when the economic driving factors are negative. This also shows that when the economic driving factor is negative, according to the concept of the decoupling effort model, the decoupling state between environmental pressure and economic benefits after eliminating economic factors is explored. Scholars Diakoulaki and Mandaraka’s decoupling effort equation in this paper expresses that when the contribution of economic factors is negative, the molecules of the total decoupling effect index formula are the state after decoupling effort and then subtract the economic factors, which is quite the same as the elimination of economic factors for environmental pressure twice. This paper lists what scholars may want to express as shown in Equation (21).

In Equations (20) and (21), D_t represents the total decoupling effect of carbon emissions. When $D_t\ge 1$, it indicates that there is a strong decoupling relationship; when $0<D_t<1$, there is a weak decoupling relationship between the two; when $D_t\le 0$ indicates that there is no decoupling relationship between the two. The decoupling effort model is further decomposed to obtain the decoupling effect of various influencing factors, where $D_{X_2}$, $D_{X_3}$, $D_{X_4}$, $D_{X_5}$, $D_{X_6}$, $D_{X_8}$, $D_{X_9}$, $D_{X_{10}}$ are the corresponding influencing factors X₂, X₃, X₄, X₅, X₆, X₈, X₉, X₁₀ decoupling effect.

4. Result Analysis

4.1. Analysis of Power Carbon Emission Characteristics

This paper divides the evolution of power carbon emissions in the Yangtze River Economic Belt into three stages: The first stage represents the rapid growth stage: from 2000 to 2011, the carbon emissions generated by the power industry increased by 553 million tons from 2000 to 2007. During the period of rapid industrial development and urbanization the construction process accelerated, driving the increase in electricity demand. From 2007 to 2009, the growth of carbon emissions from electricity was not significant, or even regressive, which may be due to the Yangtze River Economic Belt in response to 2007 ‘speed up the shutdown of small thermal power units’ policy, and actively implement the shutdown of small thermal power units. Since then, carbon emissions have a clear upward trend. The second stage represents the fluctuation stage: from 2011 to 2015, the carbon emission showed a fluctuation of decline–rise–decline. The Air Pollutant Emission Standards for Thermal Power Plants, updated in 2011, indirectly improved the efficiency of thermal power generation and reduced the emission of thermal power pollutants. In addition, hydropower and other clean energy technologies continue to develop and build, to make up for some of the electricity demand. After carbon emissions peaked in 2013, from 2013–2015 the carbon emissions of the power industry fell back. The third stage represents the small increase stage: from 2015 to 2019, carbon emissions gradually increased. As the economy of the Yangtze River Economic Belt enters the new normal, thermal power supply has been expanded to ensure the safety of power supply in the Yangtze River Economic Belt. During this period the Yangtze River Economic Belt power carbon emissions increased 216 million tons. Carbon emissions from electricity fell slightly from 2019–2020 under the impact of the pandemic (Figure 1)

The lower reaches power carbon emissions gradually increased from 2000 to 2013, reaching a phased peak of 761 million tons in 2013 (Figure 1, Figure 2, Figure 3 and Figure 4 and

Table 5. Upper, Middle and Lower Reaches of Yangtze River Economic Belt.

Regions	Provincal Level
upper reaches	Chongqing, Sichuan, Guizhou, Yunnan
middle reaches	Hubei, Hunan, Jiangxi
lower reaches	Shanghai, Jiangsu, Zhejiang, Anhui

). Since then, carbon emissions from electricity have gradually increased from 2014 to 2018. The carbon emissions of electricity in the middle reaches of the region showed an increasing trend from 2000 to 2011, reaching 240 million tons in 2011. Carbon emissions from middle reaches electricity fell by 30 million tons from 2019–2020 as the pandemic hit, the sharpest drop among the three reaches. Upper reaches power carbon emissions maintained a rapid growth trend from 2000 to 2007. Since then, the growth trend of carbon emissions has slowed down, reaching 270 million tons in 2011. From 2013 to 2016 the total carbon emissions decreased significantly, and then they increased. In 2020, the carbon emissions of electricity in the upper, middle and lower reaches of the Yangtze River Economic Belt accounted for 18.58%, 20.09% and 61.33% of the total carbon emissions of the Yangtze River Economic Belt. The lower reaches are always the main contribution area of power carbon emissions. From 2003 to 2011 the proportion of lower reaches power carbon emissions in the total power carbon emissions of the Yangtze River Economic Belt was less than 60%, while in 2012 the proportion exceeded and remained above 60%. The proportion of power carbon emissions in the middle reaches exceeded 20% for the first time in 2018, remaining above 20% since then. In 2000, China began to implement the western development strategy. The proportion of upper reaches carbon emissions increased from 20.88% in 2000 to 26.29% in 2006. The implementation of the western development promotes the growth of economic and power carbon emissions. The upper reaches accounted for less than 20% for the first time in 2015, and the proportion has remained below 20% since then. In 2016, the Yangtze River Economic Belt Development Plan Outline proposed to guide the orderly transfer of industries in the lower reaches to the middle and upper reaches, and the middle reaches adjacent to the lower reaches to undertake more industrial transfers from the lower reaches; this may be one of the reasons for the increase in the proportion of carbon emissions in the middle reaches. The total amount of carbon emissions from electricity in the Yangtze River Economic Belt increased steadily during the four periods of 2000–2005, 2006–2010, 2011–2015 and 2016–2020, and the spatial distribution of high-value areas of carbon emissions from electricity has a tendency to migrate eastward (Figure 5, Figure 6, Figure 7 and Figure 8).

The three provincial levels with the highest average power carbon emissions from 2000 to 2020 are Jiangsu Province, Zhejiang Province and Anhui Province, with 250.51 million tons, 136.28 million tons and 118.47 million tons, respectively. At least three provincial levels are Chongqing, Yunnan and Sichuan, with emissions of 30.06 million tons, 43.26 million tons and 49.63 million tons, respectively. On the one hand, due to the developed lower reaches economy, numerous industries and population agglomeration, the total carbon emissions of lower reaches are significantly higher than those of upper and middle reaches. On the other hand, the resources are significantly different, and the upper and middle reaches of the Yangtze River Economic Belt are rich in hydropower. Hydropower construction supplies regional electricity demand, thereby reducing the pressure on thermal power supply. Regarding changes in total provincial level power carbon emissions, Jiangsu Province and Anhui Province changed the most, with an increase of 256 million tons and 171 million tons, respectively, compared with 2000. Shanghai and Sichuan have the least changes, which are 18 million tons and 19 million tons, respectively, (Figure 9).

4.2. Contribution Analysis of Influencing Factors

The carbon emissions of electricity in the Yangtze River Economic Belt increased by 353.96 million tons, 352.24 million tons, 50.22 million tons and 151.56 million tons, respectively, in 2000–2005, 2005–2010, 2010–2015 and 2015–2020. During the period from 2010–2015 the incremental carbon emissions of electricity decreased significantly, and the carbon emissions of electricity increased from 2015–2020. In the lower reaches, the increment of power carbon emissions decreased from 2005 to 2010, 2010 to 2015 and 2015 to 2020, but the increment of power carbon emissions was greater than that in the middle and upper reaches. Since the implementation of the strategy for the rise of central China in 2004, the increment of carbon emissions in the middle reaches increased slightly from 2005 to 2010. In the upper reaches, the increment of power carbon emissions showed a negative growth from 2010–2015 and a positive growth from 2015–2020.

4.2.1. Absolute Factor Analysis

This paper discusses the contribution of influencing factors to carbon emissions from electricity in four stages: 2000–2005, 2005–2010, 2010–2015 and 2015–2020. The contribution rate of influencing factors in the upper, middle and lower reaches of the Yangtze River Economic Belt are shown in Figure 10, Figure 11, Figure 12 and Figure 13.

4.2.2. Absolute Factor Analysis

From 2000–2005 the economic scale of the Yangtze River Economic Belt contributed 70.36 million tons of power carbon emissions (Figure 14, Figure 15, Figure 16 and Figure 17). The contribution of economic scale to carbon emissions increased continuously from 2000 to 2005, reaching 19.9 million tons from 2014–2015. From 2000 to 2005, the proportion of thermal power in the power production structure of the Yangtze River Economic Belt reached more than 70%. With the rapid development of various industries, the demand for electricity has increased, and the economic scale and power carbon emissions have maintained simultaneous growth. The cumulative contribution of economic scale from 2005–2010 was 142.24 million tons, and the economic scale continues to drive the increase in carbon emissions. Regional GDP growth from 2007–2008 was at its lowest level from 2005–2010, during which the contribution of economic scale was below 3% (

Table 6. Contribution rate of influencing factors of power carbon emissions in the Yangtze River Economic Belt.

Year	G	C/G	E	C/E	P	C/P	T	C/T	G/P	E/G
2000–2001	0.0245	0.0024	0.0257	0.0009	0.0007	0.0264	0.0322	−0.0058	−0.0012	0.0000
2001–2002	0.0272	−0.0065	0.0202	−0.0003	0.0013	0.0191	0.0213	−0.0014	−0.0013	0.0000
2002–2003	0.0315	0.0289	0.0626	−0.0002	0.0020	0.0603	0.0594	0.0005	−0.0021	−0.0021
2003–2004	0.0323	0.0030	0.0350	0.0000	0.0015	0.0342	0.0395	−0.0050	−0.0020	−0.0001
2004–2005	0.0327	0.0115	0.0449	−0.0002	−0.0012	0.0469	0.0370	0.0069	−0.0025	−0.0005
2005–2006	0.0350	0.0093	0.0462	−0.0016	0.0009	0.0444	0.0393	0.0044	−0.0025	−0.0004
2006–2007	0.0370	−0.0129	0.0216	0.0004	0.0011	0.0218	0.0288	−0.0067	−0.0024	−0.0001
2007–2008	0.0292	−0.0320	−0.0051	−0.0005	0.0012	−0.0071	0.0223	−0.0280	−0.0012	−0.0017
2008–2009	0.0303	−0.0128	0.0161	−0.0001	0.0012	0.0153	0.0207	−0.0046	−0.0015	−0.0001
2009–2010	0.0342	−0.0023	0.0219	0.0087	0.0009	0.0306	0.0322	−0.0017	−0.0020	−0.0001
2010–2011	0.0305	−0.0021	0.0296	−0.0018	0.0024	0.0259	0.0227	0.0049	−0.0016	0.0000
2011–2012	0.0254	−0.0292	−0.0066	0.0006	0.0017	−0.0080	0.0149	−0.0210	−0.0008	−0.0015
2012–2013	0.0250	−0.0025	0.0156	0.0062	0.0016	0.0206	0.0228	−0.0010	−0.0010	−0.0001
2013–2014	0.0210	−0.0402	−0.0195	−0.0020	0.0015	−0.0239	0.0101	−0.0322	−0.0004	−0.0029
2014–2015	0.0208	−0.0253	−0.0084	0.0024	0.0012	−0.0076	0.0034	−0.0096	−0.0005	−0.0014
2015–2016	0.0203	−0.0115	0.0078	0.0001	0.0017	0.0064	0.0154	−0.0075	−0.0006	−0.0002
2016–2017	0.0200	−0.0065	0.0151	−0.0021	0.0014	0.0118	0.0143	−0.0013	−0.0007	0.0000
2017–2018	0.0189	−0.0013	0.0151	0.0022	0.0011	0.0165	0.0169	0.0004	−0.0006	0.0000
2018–2019	0.0172	−0.0096	0.0054	0.0016	0.0011	0.0061	0.0097	−0.0026	−0.0005	−0.0002
2019–2020	0.0064	−0.0180	−0.0138	0.0019	0.0000	−0.0122	0.0061	−0.0182	0.0000	−0.0008

) for the first time, contributing 26.71 million tons. From 2010 to 2015, the cumulative contribution of economic scale to carbon emissions was 141.69 million tons, and the contribution of economic scale gradually decreased to 23.87 million tons from 2014–2015. The proportion of tertiary industry in GDP increased from 42.84% in 2010 to 48.25% in 2015. The secondary industry saw the original 48.52% fall to 44.01% in 2015, with deepening economic restructuring, and the service industry development accelerated significantly. In the upper, middle and lower reaches economic scale is also the main driving factor, with a cumulative contribution rate of more than 50%. The cumulative contribution of lower reaches economic scale to power carbon emissions is 261 million tons, which is much higher than the cumulative contribution of upper and middle reaches economic scale to carbon emissions of 106.16 million tons and 84.62 million tons. The cumulative contribution of upper and lower reaches economic scale showed an upward–downward–downward state in 2000–2005, 2005–2010, 2010–2015 and 2015–2020. In 2000–2005, 2005–2010, and 2010–2015, the cumulative contribution of economic scale in the middle reaches showed a gradual increase. Overall, the contribution intensity of economic scale to power carbon emissions has weakened, but it is still the main factor driving carbon emissions. The contradiction between economic development and carbon emission reduction still exists.

From 2000 to 2005, the cumulative contribution of energy consumption in the Yangtze River Economic Belt was 90.20 million tons, and the rapid expansion of thermal power led to an increase in power energy consumption. The cumulative contribution of energy consumption from 2005–2010 was 81.66 million tons. From 2007 to 2008 the total energy consumption decreased from 347.78 to 340.49 million tons of standard coal (Figure 18), and the contribution of energy consumption was inhibited for the first time. In the 2010–2015 period, with the promulgation and implementation of energy-saving emission reduction policies and increase support for renewable energy generation, the scale of hydropower and other clean energy power generation has expanded, and thermal power has stopped expanding rapidly. From 2013–2015, the total energy consumption gradually decreased, and the contribution of energy consumption scale slowed down. The continuous growth of social electricity consumption from 2015–2020 meant that the scale of thermal power generation expanded, driving the consumption of energy. During this period, the cumulative contribution of energy consumption was 33.23 million tons. In the four periods, the cumulative contribution of lower reaches energy consumption scale gradually decreased. The cumulative contribution of upper and middle reaches energy consumption in 2005–2010, 2010–2015 and 2015–2020 shows a downward–upward state. Energy consumption is the main positive driving factor for the increase in carbon emissions in Yangtze River Economic Belt and its regions. The reduction in energy consumption directly affects the carbon emission reduction.

In 2000–2005, 2005–2010 and 2010–2015, the population scale of the Yangtze River Economic Belt increased by 8.86 million, 12.19 million and 2.01 million, respectively. The growth of the total population will lead to the expansion of the consumption scale of residents to a certain extent. Indirect promotion of enterprises will increase capacity supply and stimulate electricity demand, thus the cumulative contribution of population size in these three periods has gradually increased. As China’s coastal areas, the lower reaches have developed rapidly, with a strong population siphon ability. In 2000–2005, 2005–2010 and 2015–2020, the cumulative contribution of lower reaches population scale to power carbon emissions exceeds the cumulative contribution of population size in the Yangtze River Economic Belt. The population scale of the middle reaches in 2015–2020 was negative for power carbon emissions, due to the impact of the pandemic. The year-end population growth rate in the middle reaches was −1.02% from 2019 to 2020. The contribution to carbon emissions from electricity in 2019–2020 was −0.69 million tons, exceeding the sum of the positive driving contribution of population size in 2015–2019.

In 2000–2005 and 2005–2010, thermal power generation accounted for more than 65% of electricity output. Electricity output and carbon emissions growth kept pace. In 2005–2010 the cumulative contribution of electricity output reached 121.63 million tons. In 2010–2015, the pulling effect of output scale declined. In 2006 the renewable energy law promulgated and implemented the power production structure to clean energy generation transformation and development. In 2015 the proportion of thermal power generation was below 60% for the first time. From 2015–2020 the scale of clean energy generation continued to expand, but in order to maintain the safety of power supply and demand thermal power scale has also increased. The cumulative contribution of output scale in this period is basically the same as that in 2010–2015. In 2000–2005, 2005–2010 and 2010–2015, the cumulative carbon emissions of the upper, middle, and lower reaches output scale to the carbon emissions showed a rising–falling state. As one of the main driving factors of power carbon emissions, the optimization of electricity output structure can further reduce the contribution of output scale to carbon emissions.

4.2.3. Relative Factor Analysis

With the rapid economic development of the Yangtze River Economic Belt from 2000 to 2005, many new thermal power units were built to meet the demand for electricity. The cumulative contribution of carbon intensity from 2000 to 2005 was 19.31 million tons. In 2005–2010, 2010–2015 and 2015–2020, the emerging industries in the Yangtze River Economic Belt have further developed, and the economic value of unit power has increased. In addition, the implementation of programs such as Energy-saving Medium and Long-term Special Planning (2004) and Controlling Greenhouse Gas Emissions (2016). Increase the research and development and investment of thermal power emission reduction technology, thus the cumulative contribution of carbon intensity to carbon emissions in 2005–2010, 2010–2015 and 2015–2020 is negative. The cumulative contribution of upper, middle and lower reaches carbon intensity from 2000–2005 was positive. Among them, the upper reaches had the most cumulative contribution. The reason is that the proportion of thermal power to total power generation increased by 5.84% during 2000–2005. The lower reaches economy is large, and the economic level is high. The inhibitory effect of power carbon intensity on the cumulative contribution of carbon emissions is greater than that of the upper and middle. Due to the rapid development of hydropower in the upper reaches, the upper reaches carbon intensity had a greater inhibitory effect on power carbon emissions in 2010–2015 and 2015–2020 than in the middle reaches.

In 2007 China issued Energy Development Plan, and in 2010 introduced the Strengthen the elimination of backward production capacity notice. The Yangtze River Economic Belt has increased its support for clean energy power generation. By 2015, the installed capacity of hydropower in the Yangtze River Economic Belt was 225.37 million kWh, 4.7 times higher than that in 2000. Clean energy power production has gradually become an important hub of the power system. In addition, the energy consumption per unit of thermal power output has decreased and the electricity efficiency of energy conversion has improved. Compared with 2000, the standard coal consumption of thermal power supply decreased from 381.87 (g standard coal/kWh) to 301.79 (g standard coal/kWh) in 2015 (Figure 19). Furthermore, in 2000–2005, 2005–2010 and 2010–2015, the output carbon intensity inhibiting carbon emissions increased. The cumulative contribution of upper reaches output carbon intensity to carbon emissions in 2005–2010 and 2010–2015 was greater than that in the middle and lower reaches. This was due to the optimization of the power structure in the upper reaches of the Yangtze River, and the shift to non-fossil energy generation.

The change in energy consumption carbon intensity reflects the change in power energy consumption structure, the Yangtze River Economic Belt geographical conditions, and resource endowments and other reasons limit; in addition, the high cost of oil power generation, natural gas fuel and high cost of transmission and distribution links. Thus, during the study period, the consumption of electricity production oil decreased, while the consumption of natural gas increased slightly (Figure 18). Coal consumption accounted for 96.98% and 95.02% of the total energy consumption in 2000 and 2020, respectively. At the present time, the thermal electricity production in the Yangtze River Economic Belt is still dominated by coal energy consumption. Therefore, energy consumption carbon intensity has a slight pulling effect on electricity carbon emissions. The cumulative contribution of carbon intensity of energy consumption in the upper and lower reaches of the Yangtze River Economic Belt in the three periods of 2005–2010, 2010–2015 and 2015–2020 was positively driven.

The cumulative contribution of population carbon intensity in the Yangtze River Economic Belt in 2000–2005 and 2005–2010 was 89.84 million tons and 86.20 million tons, respectively. The living standards of residents in the region have improved and the level of electrification has increased. The rising per capita electricity demand of residents indirectly leads to the expansion of thermal power to meet the power supply. The cumulative contribution of population carbon intensity decreased significantly from 2010 to 2015 and increased slightly from 2015 to 2020. The cumulative contribution of lower reaches population carbon intensity showed a decreasing trend in four periods. The cumulative carbon emission contribution of upper reaches population carbon intensity from 2010–2015 was negatively driven. The cumulative contribution of the middle reaches to population carbon intensity in 2010–2015 was the lowest in its four periods. To further realize the carbon emission reduction, it is necessary to raise residents’ awareness of energy-saving electricity and promote sustainable development among economy, environment and population.

Per capita GDP reflects the level of regional economic development. Objectively, the increase in per capita GDP in the early stage of economic development will promote power carbon emissions, while in the GDIM method, per capita GDP has a weak inhibitory effect on carbon emissions. Vaninsky points out that per capita GDP is correlated with several factors, which are affected by Equation (13). Therefore, the impact of per capita GDP changes on power carbon emissions is only partially attributed to it. The other part is included in the impact of other factors on the carbon emissions such as Equation (15), which may lead to the inhibition of per capita GDP. On the other hand, the two relative factors of per capita GDP and energy intensity are composed of three absolute variables: they are economic, energy consumption and population. Among them, the absolute variables of economic scale are the molecules and denominators of per capita GDP and energy intensity factors, respectively. The absolute factors economy in per capita GDP is distributed not only in the previous economic scale and carbon intensity, but also in per capita GDP and energy intensity. Therefore, in the per capita GDP, the population scale may have a higher degree of carbonization of carbon emissions and play a reverse role as a denominator, resulting in a weak inhibition of per capita GDP. The above reasons may also be the reason why the cumulative contribution of per capita GDP and energy intensity to power carbon emissions is relatively lower than other influencing factors.

During the study period, the power energy consumption of the Yangtze River Economic Belt supported the economic growth rate of 10.22% with an average annual growth rate of 6.16%. With the progress of power technology, power energy efficiency has been improved and power energy consumption caused by unit GDP has gradually declined. Therefore, in the three periods from 2000–2005, 2005–2010 and 2010–2015, the inhibition of energy intensity on carbon emissions increased. The cumulative contribution of energy intensity in the four periods of upper, middle and lower was negative. The upper, middle and lower reaches in 2000–2005, 2005–2010, 2010–2015 energy intensity cumulative negative driving effect of carbon emissions increased. In 2015–2020, the cumulative contribution of upper, middle and lower reaches energy intensity negative driving weakened.

4.3. Analysis on Decoupling Effect of Electricity Carbon Emission

4.3.1. Analysis of Total Decoupling Effect Index

This paper calculates the total decoupling effect of power carbon emissions and the decoupling effect of influencing factors, with 2000 as the base period. During the study period, the power carbon emissions in the Yangtze River Economic Belt did not achieve decoupling (

Table 7. The decoupling effect of power carbon emissions in the Yangtze River Economic Belt.

Year	${\mathit{D}}_{{\mathit{X}}_2}$	${\mathit{D}}_{{\mathit{X}}_3}$	${\mathit{D}}_{{\mathit{X}}_4}$	${\mathit{D}}_{{\mathit{X}}_5}$	${\mathit{D}}_{{\mathit{X}}_6}$	${\mathit{D}}_{{\mathit{X}}_8}$	${\mathit{D}}_{{\mathit{X}}_9}$	${\mathit{D}}_{{\mathit{X}}_{10}}$	Dt
2001	−0.042	−0.454	−0.016	−0.012	−0.467	0.102	0.021	0.000	−0.867
2002	0.043	−0.436	−0.005	−0.019	−0.432	0.066	0.024	0.000	−0.758
2003	−0.136	−0.561	−0.001	−0.020	−0.546	0.031	0.023	0.011	−1.200
2004	−0.105	−0.537	−0.001	−0.020	−0.524	0.043	0.025	0.008	−1.111
2005	−0.121	−0.565	0.000	−0.010	−0.563	0.006	0.028	0.008	−1.219
2006	−0.122	−0.579	0.005	−0.011	−0.572	−0.010	0.029	0.007	−1.253
2007	−0.056	−0.528	0.003	−0.012	−0.522	0.013	0.031	0.006	−1.066
2008	0.045	−0.434	0.004	−0.014	−0.424	0.092	0.029	0.010	−0.691
2009	0.071	−0.419	0.004	−0.015	−0.408	0.092	0.030	0.009	−0.637
2010	0.066	−0.406	−0.016	−0.015	−0.416	0.082	0.030	0.008	−0.668
2011	0.063	−0.424	−0.010	−0.019	−0.424	0.061	0.030	0.007	−0.716
2012	0.122	−0.371	−0.011	−0.021	−0.368	0.103	0.029	0.010	−0.506
2013	0.116	−0.366	−0.022	−0.022	−0.374	0.095	0.028	0.009	−0.536
2014	0.190	−0.304	−0.017	−0.024	−0.302	0.154	0.027	0.014	−0.261
2015	0.226	−0.276	−0.020	−0.025	−0.276	0.164	0.027	0.016	−0.164
2016	0.232	−0.273	−0.019	−0.026	−0.270	0.167	0.026	0.015	−0.148
2017	0.229	−0.282	−0.015	−0.027	−0.274	0.160	0.026	0.015	−0.168
2018	0.219	−0.290	−0.017	−0.027	−0.285	0.150	0.025	0.014	−0.211
2019	0.224	−0.286	−0.019	−0.028	−0.282	0.148	0.025	0.013	−0.204
2020	0.249	−0.258	−0.022	−0.027	−0.257	0.174	0.025	0.014	−0.102

). This was because the thermal power generation in the Yangtze River Economic Belt was still the main position, and the economic development needs more power supply, thus the economic benefits and environmental pressures have not been decoupled. With the efforts made by the power industry and the government, the degree of non-decoupling of the total decoupling effect index has narrowed. From 2001 to 2005, power generation was dominated by thermal power, and clean energy power generation was in its infancy. The increase in power generation led to many carbon emissions, with the total decoupling effect index at a low value. Strengthen energy conservation work (2006), shut down small thermal power units (2007) and other policy implementation. Effectively control and inhibit the rapid growth of power carbon emissions, thus the total decoupling effect showed a rapid rise after 2006. From 2011–2013 the total decoupling effect index showed a rising-falling state. The total decoupling effect index gradually increased from 2013 to 2016, reaching a phased peak in 2016. The overall decoupling effect index declined in 2016–2018. The average value of the total decoupling effect index of the Yangtze River Economic Belt from 2016–2020 was slightly higher than that in 2011–2015. However, due to the steady economic growth of the Yangtze River Economic Belt, thermal power plays a supporting role in supply. During the period 2016–2020, the total decoupling effect index did not show an upward trend but changed in a certain range. It showed that the decoupling between environmental pressure and economic benefits has entered a pressure period (Figure 20).

The total decoupling effect index in the early stage of lower reaches was higher than that in the middle and upper reaches. It shows that the decoupling state between lower reaches environmental pressure and economic benefit is relatively better than that of upper and middle reaches. In 2020, the total decoupling effect index of upper, middle and lower reaches increased by 1.58, 0.948 and 0.395, respectively, compared with 2001. The upper and middle reaches power industries and the power emission reduction measures taken by the government have more room for progress. One of the main factors is hydropower as China’s second largest power generation. During the study period, hydropower in the upper and middle reaches achieved rapid development, which made up for the expansion of some power demand. The total decoupling effect index of power carbon emissions in the upper reaches of the Yangtze River Economic Belt in 2014 showed a decoupling state. Since then, the total decoupling effect has continued to maintain a decoupling state, but the total decoupling effect index has not been further improved. The total decoupling effect index in the middle reaches of the Yangtze River Economic Belt increased by more than 0.5 in 2004–2005, 2007–2008 and 2011–2012. However, the total decoupling index fluctuated in a certain range during 2012–2020, and there was no obvious upward trend.

4.3.2. Decoupling Effect Index Analysis of Influencing Factors

The decoupling effect of energy consumption scale is generally consistent with the change trend of total decoupling effect index. Due to the limitation of China’s resource endowment, the transformation of thermal power consumption energy varieties in the Yangtze River Economic Belt has limitations, and energy consumption hinders the decoupling of power carbon emissions in the Yangtze River Economic Belt. Carbon intensity decoupling of energy consumption has a weak inhibitory effect on the total decoupling effect. As the main body of thermal power consumption has not changed, the decoupling effect of energy consumption carbon intensity inhibits the realization of power carbon emission decoupling. The Yangtze River Economic Belt population grew, increasing demand for electricity and requiring the power sector to expand production capacity. At the same time, people‘s living standards continued to improve, and per capita electricity demand also increased. Thermal power plays a major role in power supply. Therefore, population size and population carbon intensity also inhibit the decoupling of electricity carbon emissions.

With the continuous economic growth of the Yangtze River Economic Belt, the power generation level and energy efficiency utilization rate have been improved and broken through. In addition, clean energy power generation with hydropower as the main body has achieved rapid expansion and construction, and power production and installation structure have been optimized. Therefore, the decoupling effect of carbon intensity and output carbon intensity play a major role in promoting the decoupling of power carbon emissions in the Yangtze River Economic Belt. The growth rate of energy consumption is lower than that of GDP, and the unit GDP produces less power energy consumption. The energy intensity also promotes the decoupling of carbon emissions in the Yangtze River Economic Belt. The decoupling effect of carbon intensity in 2020 was the largest decoupling effect in the lower reaches and the middle reaches of the Yangtze River Economic Belt, and the decoupling effect of output carbon intensity is the largest decoupling effect in the upper reaches of the Yangtze River Economic Belt. Population carbon intensity and energy consumption are the biggest factors inhibiting decoupling in the upper, middle and lower reaches of the Yangtze River Economic Belt. It can be seen that residents’ consumption will drive the demand for electricity and increase the embodied carbon emissions. Therefore, actively promoting the upgrading of residents’ consumption structure and encouraging and guiding residents’ green consumption and energy-saving electricity consumption are the main ways to achieve decoupling of carbon emissions.

5. Conclusions and Recommendations

5.1. Conclusions

In this paper, the GDIM model is extended, and the selected indicators of influencing factors of power carbon emissions basically cover the research of scholars in the power GDIM model. The contribution of influencing factors of carbon emissions in this paper is generally consistent with the existing research results in the final driving direction. In this study, due to the selection of more comprehensive indicators of carbon emission influencing factors, the distribution of power carbon emissions is reflected in a more systematic indicator system, therefore the analysis of the influencing factors of power carbon emissions is more objective. The following conclusions are obtained through research:

The increment of carbon emissions in the Yangtze River Economic Belt was basically the same in 2000–2015 and 2005–2010, and the increment of carbon emissions from electricity increased slightly in 2015–2020. The total carbon emissions of electricity in the three major regions of the Yangtze River Economic Belt are significantly different. The lower reaches are always the main contribution area of electricity carbon emissions. The total power carbon emissions in the upper and middle reaches are relatively similar. Overall, the proportion of upper and middle reaches power carbon emissions to the total carbon emissions of the Yangtze River Economic Belt have slightly decreased and increased, respectively. The high value of power carbon emissions in the Yangtze River Economic Belt has an eastward migration trend.

Economic scale is the largest positive driving factor of power carbon emissions in the middle and lower reaches, while carbon intensity is the largest negative driving factor. Electricity output is the largest positive driver of upper reaches carbon emissions, while output carbon intensity is the largest negative driving factor. In 2000–2005, 2005–2010 and 2010–2015, the output carbon intensity of the Yangtze River Economic Belt gradually increased the inhibition of power carbon emissions. The cumulative contribution of the lower reaches population scale to power carbon emissions in 2000–2005, 2005–2010 and 2015–2020 exceeded the cumulative contribution of population size in the Yangtze River Economic Belt. In 2000–2005, 2005–2010, 2010–2015 and 2015–2020, the cumulative contribution of upper and lower reaches economic scale showed an upward–downward–downward state. In 2000–2005, 2005–2010 and 2010–2015 the energy intensity of the upper, middle and lower reaches gradually increased the inhibition of power carbon emissions.

The total decoupling effect index of power carbon emissions in the Yangtze River Economic Belt in 2006–2008 and 2011–2015 has improved significantly. The decoupling effect of energy consumption and population scale carbon intensity has a major inhibitory effect on the decoupling of power carbon emissions. The decoupling effect of carbon intensity and output carbon intensity plays a key role in promoting the total decoupling effect of power carbon emissions. The upper reaches of the Yangtze River Economic Belt achieved decoupling in 2014–2020. During 2016–2020, the total decoupling effect index of the upper, middle and lower reaches of the Yangtze River Economic Belt fluctuated within a certain range. The decoupling between environmental pressure and economic benefits in the power industry has entered a certain pressure period.

5.2. Recommendations

The clean and low-carbon transformation of power in the Yangtze River Economic Belt has played an important role in China’s power peaking. As the economy returns to steady development in the post-epidemic era, the pressure on power carbon emission reduction still exists. In order to realize the green and low-carbon transformation development of the Yangtze River economic belt, the following suggestions are put forward:

The cumulative contribution rate of the economic scale of the upper, middle and lower reaches of the Yangtze River Economic Belt is more than 50%. The cumulative contribution rate of population carbon intensity in the Yangtze River Economic Belt is 32.75%. The improvement of regional economy, residents’ lives and electrification level will put pressure on power carbon emission reduction to a certain extent. Reduce the contribution of economic and demographic factors to power carbon emissions from the demand side. Actively develop the circular economy of the Yangtze River Economic Belt, accelerate the promotion of modern service industries, promote the development of emerging industries such as digital economy, and rely on technological progress and innovation to drive industrial growth. Population carbon intensity is the main factor hindering decoupling, thus strengthen energy conservation and emission reduction publicity, enhance residents’ awareness of responsibility for energy conservation and emission reduction, and promote energy conservation and low-carbon life.

During the 13th Five-Year Plan period (2016–2020), the total decoupling effect index of the Yangtze River Economic Belt entered a certain pressure period. Wind and solar power are the key to further decoupling carbon emissions from electricity, and there is a need to actively respond to the 14th Five-Year Plan to vigorously increase the scale of wind and photovoltaic power generation. In addition, scientific planning and development of hydropower, nuclear power and other clean energy generation to make up for the withdrawal of backward thermal power supply. The cumulative contribution rate of electricity output in the upper, middle and lower reaches of the Yangtze River Economic Belt is more than 40%. From the supply side to slow down the positive driving of output factors on power carbon emissions. Continue to eliminate backward thermal power capacity and increase the use of high-efficiency ultra-supercritical generating units. The use of advanced technology carbon capture, utilization and storage to improve the economic and environmental benefits of coal-fired power plants. In addition, increase investment in research and development in photovoltaic, wind power and other fields to achieve technological breakthroughs and reduce construction costs. Adhere to the development of centralized and decentralized clean energy power generation, continue to strengthen the construction of new energy power consumption and cross-regional transmission capacity, and achieve efficient operation of clean energy power generation.

The cumulative contribution rate of power carbon intensity in the Yangtze River Economic Belt is −15.76%, and the cumulative contribution rate of output carbon intensity is −12.93%. However, the cumulative contribution rate of energy consumption in the upper, middle and lower reaches of the Yangtze River Economic Belt is more than 30%. Power production structure optimization can effectively improve the carbon intensity and output carbon intensity factors to curb carbon emissions. At the same time, the promotional effect of energy consumption on power carbon emissions is reduced. Therefore, adhere to local conditions to develop clean energy. The upper reaches still have the potential for hydropower development, requiring scientific planning and construction of hydropower projects in upper reaches Yunnan and Sichuan. Accelerate the construction of upper reaches hydropower bases and transmission channels and expand the scale of power transmission to lower reaches. Middle reaches combined with geographical conditions to develop wind power, photovoltaic power generation, play middle reaches water supply advantages to make up for the growth in demand for electricity. Further play to the geographical advantages of lower reaches coastal, make full use of Zhejiang, Jiangsu coastal wind energy advantage construction of wind power, and the safe development of coastal advanced nuclear power construction.