Unraveling the Surrounding Drivers of Interprovincial Trade Embodied Energy Flow Based on the MRIO Model: A Case Study in China

Abstract

To achieve the carbon neutrality target, China has proposed “dual control” policies on provincial energy consumption. However, inter-provincial trade drives significant embodied energy flows beyond local demand. How do we identify key energy consumers driving through other provinces? And how does energy, especially from coal, flow to other provinces? Current studies analyzed regional and sectoral energy flow, which are always separated. And seldom was attention paid to coal flow. Intending to identify the critical energy-consuming province in China and investigate how energy and coal flow out from it, this study applied the EE-MRIO model to measure energy and coal embodied in provincial trades. The results suggest the following: (1) The energy embodied in provincial trade was mostly from energy-rich regions to provinces that lacked energy but had developed economies. Shanxi is a critical embodied-energy export province; (2) neighboring provinces and economically developed provinces drive the most embodied energy from Shanxi, and embodied energy mainly flows from the energy sectors and high-energy-intensity sectors; and (3) the provincial and sectoral coal flow in Shanxi presents consistent characteristics of embodied energy flow. We contributed to understanding the energy equity affected by embodied energy flow and propose energy consumption as a relieving measure.

1. Introduction

To realize affordable and clean energy as mandated by the United Nations Sustainable Development Goals (SDG) 7 [1], nations take steps to lower energy intensity and reduce energy consumption. China, as the world’s largest energy consumer [2], whose total energy consumption has increased by nearly 23% to 5.72 billion tons of standard coal from 2013 to 2023 [3], faces the challenge of an unbalanced energy structure that is over-dependent on fossil energy, with coal accounting for more than 60% [4]. Aimed at relieving energy pressure, China has adopted a series of “dual control” policies to regulate total energy consumption and intensity [5]. However, in regional trade, production-side sectors embed embodied energy into products flowing to consumer sectors. However, territory-based green policies place more decarbonization burdens on producers. The spatial mismatch of energy consumption and decarbonization missions highlights the regional energy equity issue [6]. Thus, it is important to identify the critical energy consumption provinces and investigate how energy and coal flow in these provinces.

Energy flows pervert the energy equity problem, of which energy compensation is a solution. Through industrial chains, energy embodied in regional trades flows from energy-intensive producers to industrialized consumers. That means, production-side sectors consume energy not only for local demands but also for industrialized regions’ demands [7]. However, existing energy policies assign emission responsibilities based on producers’ direct energy use, causing energy-intensive regions to pay the environmental price for developed regions’ consumption [8], highlighting the energy equity issue. Energy equity is a subsidiary issue of environmental equity, implying fair access to energy resources, with no one asked to bear a greater energy burden [9]. In this study, we are referring to regional energy equity [10,11]. Many eco-compensation tools were put forward to solve environmental equity issues, such as carbon compensation [12] and energy compensation. Energy compensation mechanisms provide compensation for regions that are adversely affected by energy exploration and consumption, asking benefited regions to pay extra environmental compensation to product providers, which can eliminate the free-ride phenomenon [13].

Previous studies have investigated resource-rich regions providing for other provinces’ energy needs. (1) Identifying the energy exporter is the first thing to do before analyzing the energy flows. Production-based, consumption-based, and income-based are the three main accounting principles. Scholars often use one of the accounting principles to assess environmental issues (e.g., energy consumption, carbon emissions, and other pollution) in regions. Some scholars also paid attention to clarifying the differences between different accounting principles. (2) Embodied energy flows between regions and sectors. At a regional level, scholars classified regions into energy exporters, importers, and intermediate transmitters. They investigated how embodied energy inflows and outflows from regions. At a sectoral level, the embodied energy flows of energy-intensive sectors received mass attention. (3) Coal is a main fuel, and coal flow plays an important role in coal consumption. Like the energy-flowing trend, coal flows from developed regions to less-developed regions. Developed countries import coal-intensive products from foreign countries, and developing countries consume coal to satisfy foreign countries’ needs.

However, existing studies suffer from shortcomings in several ways: (1) According to different accounting principles, a region’s energy consumption is different. Combining consumption-based and income-based energy consumption, we can better comprehend what role a region plays in the industrial energy-consumption chain. (2) The regional analyses and sectoral analyses are always separated. Learning about the energy flows from particular sectors in particular regions can help us make more precise energy conservation policies. (3) Existing studies were mostly conducted at the global level. Since energy conservation policies are always processed nationwide, a smaller-scale analysis is needed.

Our study is aimed at investigating how resource-rich provinces satisfy other provinces’ demand for energy. In more detail, we address the following questions: (1) How to identify a key energy-consuming region? (2) To which province did its energy outflow to? From which sector did its energy outflow? (3) As coal is a dominant energy source, which province did its coal outflow to? From which sector did its coal outflow? The remainder of this article is organized as shown in Figure 1.

2. Literature Review

2.1. Identifying the Key Regions of Energy Consumption

Unbalanced development among regions presents different natural and social conditions; consequently, their responsibilities and capabilities for energy saving vary [14]. Different natural and social conditions cause energy flows. Regarding the natural conditions, regions that have abundant energy resources are usually energy suppliers, while those that lack energy are often energy demanders. Regarding the social conditions, regions that have well-developed industries often need more energy. Production structure also has an impact on energy flow. Economically developed regions always have a sustainable industrial structure, which is dominated by tertiary sectors and has less demand for energy. Other developing regions’ industrial structure is dominated by the secondary sectors and often needs more energy. Industrial chains and inter-regional trades connect different regions, causing energy flows. Accompanied by energy flows, some regions act as energy suppliers on a fixed basis, and others as energy demanders. China’s economic structural reform has experienced a change from the demand side to the supply side, so the analysis of energy saving should also be considered from the supply side and the consumption-side. Energy saving depends on identifying critical energy users and needs to account for the regions’ energy consumption.

There are three accounting approaches in previous studies: production-based accounting (PBA), consumption-based accounting (CBA), and income-based accounting (IBA). PBA measures environmental issues caused by producing activities, and is widely used in international climate cooperations (e.g., the Kyoto Protocol) to constrain producers’ carbon-emitting behaviors [15]. However, with the consideration of industrial linkages, PBA cannot separate environmental responsibilities entirely. Trade redistributions geographically separate the producers from consumers and investors. For example, energy consumption in Region A may aim to meet the product demand of Region B (CBA), while the input of Region A may also lead to energy consumption in Region C (IBA). CBA covers upstream environmental impacts, allocating environmental issues to consumer activities [16]. CBA suffered from doubts because the producer’s acquisition of fortune also affects environmental issues, and it is inappropriate to let consumers take all of the responsibility [17]. IBA is complementary to CBA, which estimates the environmental impact of input on the production process [18]. In brief, CBA and IBA can, respectively, quantify the environmental responsibilities of the consumption side and supply side in industrial chains [19].

2.2. Energy Flow

Existing studies investigated how energy flowed between regions through the industrial chains. From a worldwide perspective, with the development of global value chains, more production activities not only simply meet local needs but also satisfy the demands of other countries [20]. Developing countries consume large amounts of energy due to their expansion of economic growth, among which productions and services from developed countries also take a share. Meanwhile, according to the pollution haven hypothesis, developed countries outsourced their production activities, which also caused energy embodied in global trade [21]. With trade protection restricting energy-intensive sectors transferred from developed countries to developing ones, domestic energy consumption in emerging developing countries decreased, while it increased in developed countries. Different economic- and trade-developing models cause various energy-consuming characteristics. The distribution of embodied energy is attributed to its role in the global value chain [22]. Countries may play the role of embodied energy exports, intermediate transmitters, and importers. For example, China’s domestic demand accounts for a large amount of embodied energy, and critical sectors in the energy-flowing paths are all from China itself [23]. From a national perspective, domestic trade also promoted energy flows. Embodied energy from northern China to the south alleviates energy pressure [24]. In addition, some studies were analyzed on a smaller scale, such as the comparison of several megacities. Considerable gaps exist in direct and embodied energy uses of megacities, shifting negative externalities inherent in environmental issues to others [25]. Regarding the unfairness caused by energy flows, existing studies have focused on regional energy policies and have either simulated or verified their effectiveness. Regional integration policy significantly improves energy efficiency in the non-resource-based cities, as well as in small- and medium-sized cities [26]. The establishment of the National Industrial Relocation Demonstration Zones has reduced energy efficiency [27]. The effectiveness of regional energy policies is uncertain, placing demands on understanding the current status of regional energy flows.

Existing studies also investigated how energy is redistributed in sectors across the industrial chains. Some existing studies have figured out key sectors in embodied energy flows. Key sectors include different kinds: with a large-scale impact, strong impact, strong intermediary ability, and strong central ability [28]. The key sectors also varied in countries. In developing countries, key sectors are always in heavy industries; in developed countries, key sectors are always in technological industries [29]. Some previous studies focused on one particular sector. The analyzed sector is either an energy-intensive sector [30] or has a close industrial linkage with other sectors [31]. Construction is the most analyzed sector, and energy embodied in buildings during their life cycles is a hot topic [32]. Energy-saving technologies and cleaner production could realize the reduction potential of the construction sector’s energy consumption [33]. In the information era, new technologies play a more and more essential role in economic development. Their development relies on electricity, which consumes large amounts of energy in its production process. So, the energy flow of new-technology sectors should also be assessed. Information and communication technology sectors’ embodied energy consumption is unignorable, and its biggest consumer is communication equipment [34].

However, regional analysis and sectoral analysis are always separated, which may generate an inaccurate consequence. Owing to different economic, social, and natural conditions, key sectors in different regions may not always be consistent in general. Critical sectors in energy consumption varied in developing countries and developed countries; the difference comes from different economic developing modes. Likewise, different regions always have different critical energy-consuming sectors. We combine these two analyses, investigating how energy outflows from energy-providing provinces and from which sector.

2.3. Coal Flow

While existing studies have mapped interprovincial carbon transfers, the role of coal as the dominant energy carrier remains understudied. Prevalent studies themed on coal consumption mostly concentrated on direct coal use. Given the background of studies related to energy consumption being widely conducted, it is vital to shift attention to coal consumption. Excavated and widely used since the Industrial Revolution, coal remained the dominant fuel and expanded its market continuously. Coal demand grew by 1.2% in 2022 and reached its highest at 8 billion tons, and is predicted to continuously increase in the future [35]. Developing countries, led by China and India, consumed more coal resources to accelerate their economic developing speed, and were required to bear more energy-saving responsibilities [36]. Many energy conservation steps were taken and received significant feedback. In China, the export structure transferred from energy-intensive to technology-intensive [4], declining energy intensity, and the Chinese government’s effort to promote renewable energy all caused a reduction in coal consumption. However, many developed nations imported coal-based commodities from less-developed countries to decrease domestic coal use. The transfer of environmental pressure misled the rational allocation of coal conservation responsibilities among nations. An insight into indirect coal use is necessary.

Some of the existing literature assesses embodied coal flow at a global level. The global coal utilization network was determined by interlinked international trade and geographic location. Mainland China is a net coal exporter, which means its coal consumption mainly meets external needs rather than local final demands [37]. On the other hand, America exploited 2/5 of the coal from foreign countries [38]. The worldwide coal flow presented a common direction: import from developing countries and export to developed countries, of which the UK is an example [39]. Coal flow can also provide new insight into how coal consumption declined. Coal consumption declined in some nations (i.e., China) because of transfers to others, while in some other countries (i.e., the UK) it was because of energy transition [39].

Previous studies investigated coal flow at a global scale; however, it is necessary to trace the coal flow at a smaller scale. The scale and technology of coal production have great differences in provincial China. Economic structure and key industries resulted in different demands and supply for coal resources, and initial coal endowment was also distributed unequally. Consequently, coal flow in China is also an essential phenomenon. In this study, we trace the coal flow in China at a provincial level, trying to put forward detailed coal compensation suggestions.

3. Methodology

3.1. Case Study

Owing to different energy resource endowments, regional energy consumption in China presents different characteristics. Provincial analysis can reflect how embodied energy flow impacts regional energy equity. Figure 2 shows Shanxi’s location in China.

Shanxi is a critical energy producer and consumer in China. Its energy consumption structure and industrial structure affect China’s carbon emission reduction plans. While relying on the development of the coal industry for a long time, Shanxi’s industrial structure and energy consumption structure are dominated by coal. Even though the 14th Five-Year Plan specifically emphasizes the dual control of energy consumption, due to the overemphasis on economic development, the emission reduction in the coking, steel, and electric power industries in Shanxi still has little effect. Under the contradiction of maintaining economic development, ensuring energy security, and realizing dual control of energy consumption, figuring out how to realize energy transformation in Shanxi is a problem worthy of long-term study.

Shanxi is also China’s first pilot province of the energy revolution. Serving as an energy supply base, Shanxi undertakes the important responsibility of supplying energy to meet the needs of the economy and to ensure China’s energy security. In the second half of 2021, Shanxi signed mid-to-long-term coal supply guarantee contracts with 14 provinces to stabilize coal prices and maintain the safety of coal industries and supply chains [40]. In 2022, the People’s Government of Shanxi Province issued opinions to stabilize coal supply and alleviate the contradiction between coal supply and demand [41]. In 2023, the CPC Shanxi Provincial Committee emphasized enhancing Shanxi’s power-supply capacity to the Beijing-Tianjin-Hebei, guaranteeing a coal power supply through the promotion of a coal-power joint operation [42]. In the form of coal, Shanxi provides energy to the whole country and ensures all provinces’ production activities.

Existing studies estimated energy consumption and coal consumption in Shanxi. As to energy consumption in Shanxi, most studies focus on the spatial–temporal characteristics of carbon emissions, energy structure, and energy intensity [43]. Due to the abundant coal resources and coal-dominated industrial structure and energy consumption mode in Shanxi, many scholars also paid attention to coal consumption. How do we realize sustainable development in resource-based small industrial and mining cities? How do we maximize economic growth and minimize pollution regarding coal consumption? What is the future coal service demand of Shanxi Province, especially under different scenarios? Scholars conducted a series of studies.

Therefore, this paper analyzes the total amount and flow path of embodied energy and coal between Shanxi and the rest of China to clarify Shanxi’s resource consumption in the economic linkage of China.

3.2. Method

In this study, to identify critical energy consumers from both backward and forward perspectives, embodied energy is firstly assessed by consumption-based and income-based accounting. Consumption-based accounting evaluates emissions occurring upstream in the supply chain, while the income-based accounting evaluates downstream emissions [44,45]. Both of these two accounts are based on input–output analysis. Consumption-based accounting derives from the row balance relationship of the input–output table, that is, $\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{o}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t}=\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{e} \mathrm{d}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}+\mathrm{f}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{l} \mathrm{d}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}$. Income-based accounting comes from the column balance relationship of the input–output table, that is, $\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{i}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}=\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{e} \mathrm{i}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}+\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{m}\mathrm{a}\mathrm{r}\mathrm{y} \mathrm{i}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}$. The two accountings respond to the Leontief inverse matrix and the Ghosh inverse matrix, respectively, and detailed descriptions and equations are as follows. Secondly, these two accounts are used to capture Shanxi province’s consumption and flow of embodied coal.

3.2.1. Consumption-Based Embodied Energy

Input–output analysis was first advocated by Leontief [46]. Multi-regional input–output analysis can be used for estimating embodied energy flows in different regions [31].

In the MRIO model, there is an identity relationship between total output and final use, which can be stated as follows:

$$X =AX +F$$

(1)

where $X$ is total output, and $F$ is final use. So, the relationship between total output and final use can also be stated as follows:

$$X ={\left(I -A\right)}^{-1}F$$

(2)

where $I$ is an identity matrix; ${L =\left(I -A\right)}^{-1}$ is Leontief inverse, which indicates the amount of production provided by each sector when one unit of final demand is increased; $A$ is the direct input coefficient matrix, where each element $a_{ij}^{rs}$ indicates the amount of intermediate input from sector $j$ in region $s$ that sector $i \mathrm{i}\mathrm{n} \mathrm{r}\mathrm{e}\mathrm{g}\mathrm{i}\mathrm{o}\mathrm{n} r$ directly needs to produce one unit product:

$$A =\left\{a_{ij}^{rs}\right\}=\left\{{\displaystyle \raisebox{1ex}{$z_{ij}^{rs}$}\!\left/ \!\raisebox{-1ex}{$x_j^s$}\right.}\right\}\left(r,s =\mathrm{1,2},\dots ,30;i,j =\mathrm{1,2},\dots ,27\right)$$

(3)

The amount of energy needed to satisfy an economy’s final use can be measured by the consumption-based method, which is embodied energy. Consumption-side embodied energy is calculated as follows:

$$E =DLF$$

(4)

where $D$ is the energy intensity matrix. $D$ is a diagonal matrix formed by $d_n^i$, and $d_n^i$ represents the energy consumption per unit output of sector $n$ in region $i$.

3.2.2. Income-Based Embodied Energy

Income-based embodied energy can be measured by the MRIO model, too. In the MRIO model, the identity relationship between total output and primary input can be written as follows:

$$X =HX +V$$

(5)

where $X$ is the total output, and $V$ is the primary input. So, the relationship between total output and primary input can be stated as follows:

$$X ={\left(I -H\right)}^{-1}V$$

(6)

where $I$ is an identity matrix; ${G =\left(I -H\right)}^{-1}$ is Ghosh inverse, which denotes the amount of production provided by each sector when one unit of primary input is increased. There is an academic debate regarding Ghosh’s inverse matrix, with some arguing that it should be understood as a price model rather than a quantity model [47]. In this study, following the existing research practices [48,49], the Ghosh inverse matrix is only used as an accounting framework to trace economic activities and the associated energy and coal flows from the income side, without involving any analysis of “price effects”. $H$ is the direct distribution coefficient matrix, where each element $h_{ij}^{rs}$ indicates the amount of intermediate input distributed from sector $i \mathrm{i}\mathrm{n} \mathrm{r}\mathrm{e}\mathrm{g}\mathrm{i}\mathrm{o}\mathrm{n} r$ that sector $j$ in region $s$ needs to produce one unit product, which can be calculated as follows:

$$H =\left\{h_{ij}^{rs}\right\}=\left\{{\displaystyle \raisebox{1ex}{$z_{ij}^{rs}$}\!\left/ \!\raisebox{-1ex}{$x_i^r$}\right.}\right\}\left(r,s =\mathrm{1,2},\dots ,30;i,j =\mathrm{1,2},\dots ,27\right)$$

(7)

The amount of energy promoted by an economy’s primary input can be measured by the income-based method, which is also embodied energy. Income-side embodied energy is calculated as follows:

$$E =VGD$$

(8)

3.2.3. Embodied Coal

Based on the consumption-based and income-based embodied energy model, this paper extracts and combines raw coal and six by-products of coal (cleaned coal, other washed coal, briquet, coke, coke oven gas, and other coking products) from all energy sources. $C$ denotes the direct coal consumption coefficient matrix.

Similarly to embodied energy calculation, consumption-based embodied coal can be calculated as the following equation:

$$EE =CLF$$

(9)

Income-based embodied coal can be calculated as follows:

$$EE =VGC$$

(10)

3.3. Data and Data Processing

In this paper, the data used include the multi-regional input–output tables of 31 provinces in 2015 and 2020, China’s provincial energy lists (30 provinces) in 2015 and 2020, and the conversion factors from physical units to coal equivalent. (1) The input–output table of 2015 and 2020 is estimated by the bi-proportional scaling (RAS) method. This method adjusts the rows and columns of the 2017 multi-regional input–output table iteratively, to make it consistent with the 2015 and 2020 extended input–output table. According to the National Bureau of Statistics, the IO table fluctuates within a 7% range. Therefore, 0.07 was used as the standard; the iteration stops when the error is less than 0.07 [50]. The RAS method has certain limitations. Firstly, its core assumption is that changes in the technical coefficient can be explained by a unified “substitution effect” and “manufacturing effect”, which may not fully capture the sector-specific technological progress. Secondly, it assumes there are no new products or disappearing sectors, and is unable to handle emerging industries that appear in the target year or traditional industries that disappear [51]. Despite the above limitations, the RAS method remains one of the commonly used tools for updating the table due to its wide application in input–output analysis. The results of this study should be interpreted within the context of the limitations of this methodology. (2) The provincial energy lists for 2015 and 2020 are from CEADs. The conversion factors from physical units to coal equivalent are from the National Energy Statistical Yearbook 2021 [52,53]. The Multi-Regional Input–Output table of 31 provinces in mainland China (42 sectors) describes the data of economic linkages between 1302 sectors in 31 regions of China (excluding Hong Kong, Macao, and Taiwan), including 22 provinces, 4 municipalities, and 5 autonomous regions. China’s provincial energy list (30 provinces and regions) describes the energy consumption of 45 sectors in 30 regions of China (except Tibet, Hong Kong, Macao, and Taiwan) in 19 kinds of energy, including raw coal, briquet coal, coke, natural gas, crude oil, electricity, and heat.

Due to the different sector classifications of the input–output table and the provincial energy list, for the convenience of calculation, this paper combines the 31 provinces and 42 departments in the input–output table and the 30 provinces and 45 departments in the provincial energy list, and carries out sector consolidation and data processing. To be more detailed, this study adopts the principle of maintaining consistent production content for sectoral aggregation. The “ferrous metals mining and dressing” sector and “nonferrous metals mining and dressing” sector in the provincial list are merged, corresponding to the “mining and processing of metal ores” sector in the input–output table, forming the consolidated sector 4 “metals mining and pressing”. Similarly, the “transport, storage, and postal services” sector and “information transfer, software and information technology services” sector in the input–output table are merged, corresponding to the “transport, storage, postal & telecommunications services” in the provincial energy list, resulting in sector 25 “transportation and storage”, thereby ensuring data comparability and consistency for subsequent calculations. We finally obtained 30 provinces (Tibet energy data is temporarily missing) and 27 sectors. Among these, there are 1 agricultural sector, 23 industrial sectors, and 3 service sectors. The 30 provinces and 27 sectors divided in this paper are summarized in Appendix A and Appendix B, and the conversion factors from physical units to coal equivalent of 19 kinds of energy are listed in Appendix C. As the energy and coal data from Tibet were not accessible, to maintain the balance of the input–output table, we set the energy coefficient and coal coefficient for Tibet to 0 when calculating embodied energy and coal, and excluded Tibet from the final results. Since, in the IO table, Tibet has few economic linkages with other provincial sectors, it will not affect the provincial linkage conclusions. These data-processing measures did not change the provincial rankings.

4. Results and Discussion

This study first analyzes the embodied energy of China’s provinces in 2015 and 2020 from the perspectives of consumption-side and income-side, trying to clarify the status of Shanxi in China’s embodied energy. Secondly, the flow of embodied energy and embodied coal in Shanxi is measured to find the provincial and sectoral flow relationships.

4.1. Identifying Critical Energy Consumption Province

4.1.1. Total Embodied Energy Consumption

Before identifying critical energy consumption provinces in China, the amount of provincial energy consumption is analyzed first. This study begins with consumption-side embodied energy. From Figure 3, we can tell that in 2015, Guangdong, Shandong, Jiangsu, Henan, and Zhejiang were the top embodied-energy consumption provinces; In 2020, the top embodied-energy consumption provinces were Shandong, Guangdong, Zhejiang, Jiangsu and Hebei. From 2015 to 2020, except for Liaoning, Chongqing, Guangxi, Beijing, Shanghai, Jilin, Tianjin, Qinghai, and Hainan, the consumption-based embodied energy of other provinces all showed varying degrees of growth. In addition, consumption-based embodied energy showed a gradual decline from east to west and from south to north in general. Industrial production in prosperous regions such as Shandong, Guangdong, Jiangsu, and Zhejiang needed more products from other regions, thus generating a lot of consumption-side embodied energy. Ningxia, Gansu, and Hainan generated less embodied energy at the consumption side. The industrial structure of Shanghai, Tianjin, and other provinces was dominated by the tertiary industry, so most products that they demanded from other provinces were light industrial products. From 2015 to 2020, there was a small increase in the consumption-side embodied energy. Nearly one-third of the provinces (Liaoning, Chongqing, Guangxi, Beijing, Shanghai, Jilin, Tianjin, Qinghai, and Hainan) experienced a decrease in consumption-side embodied energy, while the remaining provinces remained flat or even increased, indicating that energy conservation policies still have a large space to play.

Then it comes to the composition of consumption-side embodied energy. In the input–output table, the final demand is composed of rural household consumption, urban household consumption, government consumption, fixed capital formation, and changes in inventory. Figure 4a,b present the composition-side embodied energy based on different final demand types for 2015 and 2020, respectively. From the final demand composition in Figure 4, it can be seen that, on the whole, the composition-side embodied energy of the 30 provinces was quite different, and the fixed capital formation factor was always the main component of embodied energy in each province. This factor contributed 64% to embodied energy in 2015 and 59% in 2020, showing a reduced proportion. Relying on the scale advantage of industrial agglomeration and rich labor resources, Tianjin, Shandong, and Henan have attracted a large amount of investment, so the proportion of embodied energy caused by fixed capital formation factors is large. From 2015 to 2020, the composition-side embodied energy caused by consumption factors has been increasing, among which urban household consumption drove the most significant growth, indicating that the driving effect of urban household consumption on energy use is gradually strengthening. It is worth noting that the embodied energy driven by rural household consumption in Shandong, Sichuan, Shanxi, and other provinces increased rapidly, indicating that the urban–rural gap is gradually narrowing under the rural revitalization strategy. Due to the high degree of urbanization in Beijing, Tianjin, and Shanghai, the proportion of embodied energy caused by rural household consumption is very small. However, the urbanization process in central and western provinces, such as Shanxi and Gansu, was relatively slow, and the composition-side embodied energy caused by rural household consumption accounts for a large proportion.

Income-side embodied energy also matters. Figure 5 shows that income-based embodied energy decreased from north to south. Compared with the southern provinces, the northern provinces have a stronger base of heavy industry, which requires more energy-based raw materials for production. Thus, the share of income-side energy consumption embodied in heavy industry was more significant. In 2015, Shandong, Hebei, Jiangsu, Guizhou, Liaoning, and Inner Mongolia were the top income-based energy consumption provinces; in 2020, Shandong, Hebei, Jiangsu, Guangdong, Shanxi, and Inner Mongolia were the top income-based energy consumption provinces. From 2015 to 2020, the income-side embodied energy increased in most provinces, especially Hebei, Shandong, and Jiangsu, which have high levels of industrial development, dense population, and high demand for raw materials. This also shows that China’s economic development was closely related to energy consumption. It is necessary to take measures to promote the energy revolution.

Then, the composition of income-side embodied energy is analyzed. In the input–output table, the value-added matrix consists of compensation of employees, net taxes on production, depreciation on the fixed capital, and operating surplus. Figure 6a,b, respectively, present the embodied energy composition based on primary inputs for 2015 and 2020. It can be seen from Figure 6 that among the four primary input factors, the income-side embodied energy caused by compensation of employees accounted for the largest proportion, with an average proportion of 40%. From 2015 to 2020, the embodied energy consumption caused by the operating surplus in Jiangxi, Shandong, and Heilongjiang decreased to varying degrees, while the income-side embodied energy consumption caused by the labor compensation increased. Due to the short-term stability of the tax policy and the fixed asset depreciation strategy, the variation in the income-side embodied energy caused by the net taxes on the production factor and the depreciation on the fixed capital factor is not obvious.

4.1.2. Energy Flow Embodied in Inter-Provincial Trade

Embodied energy flows with inter-provincial trades. Embodied energy outflow refers to the energy consumption amount caused by one province to provide products and services to other provinces. Embodied energy inflow refers to the energy consumption amount caused by one province to purchase products and services from other provinces [54]. The inflow provinces transfer the energy consumption responsibility to the outflow provinces, causing the outflow provinces a disadvantaged position in inter-provincial economic and trade cooperation.

Figure 7a,b, respectively, show the provincial embodied energy flows in 2015 and 2020. It can be seen from Figure 7 that Hebei, Shandong, and Inner Mongolia were provinces with the largest net embodied energy outflow in 2015. The ratio of net embodied energy outflow and local consumption in Ningxia was 168%, namely, the embodied energy of Ningxia caused by inter-provincial consumption was much higher than local consumption. Guangdong, Zhejiang, Beijing, and Chongqing were provinces with the largest net embodied energy imports. The ratio of net embodied energy imports and local consumption in these provinces was higher than 100%, especially since the ratio of Beijing was 270%. The high ratio shows that inter-provincial embodied energy caused by these provinces’ consumption was more than local embodied energy consumption. The production of various sectors in these provinces depended on other provinces to a large extent. In 2020, Hebei, Inner Mongolia, and Liaoning were the top net embodied energy export provinces. The ratio of the net embodied energy outflow and local embodied energy consumption of Liaoning and Inner Mongolia was 123% and 112%, which still exceeded 100%. Guangdong, Zhejiang, Chongqing, and Beijing were still provinces with the largest net embodied energy imports, and the ratio of net embodied energy imports and local embodied energy consumption of the four provinces has increased. Overall, Hebei, Liaoning, Inner Mongolia, and other energy export provinces are rich in resources. The embodied energy consumption of these provinces is significant due to their contribution to other provinces in the inter-provincial trade. Guangdong, Zhejiang, Chongqing, Beijing, and other embodied energy import provinces have developed manufacturing industries and dense populations, but have poor resources, which makes it difficult to meet their own product demand. They have to rely on a large amount of embodied energy transferred from other provinces.

Specifically regarding the flow patterns, energy flows can be divided into two categories: those driven by geographical proximity and those driven by cross-regional industrial demands. The high-frequency flows between neighboring provinces are caused by geographical proximity, while the flows between energy-rich provinces in the central and western regions and distant coastal developed provinces mainly reflect cross-regional industrial demand-driven factors. Take the 2015 income-based energy of Shanxi, for example, about 5.5 million tons of standard coal energy flowed to neighboring provinces (Hebei, Henan, Shaanxi, and Inner Mongolia), and about 9.8 million tons of standard coal energy flowed to developed provinces (Shanghai, Jiangsu, Zhejiang, and Shandong). Both of these two energy flow patterns have significant impacts on the energy consumption in Shanxi Province.

Inner-provincial trade accounts for a large proportion of embodied energy consumption. To clarify the embodied energy flows between provinces, only the top 100 inner-provincial embodied energy flows and the embodied energy flows are presented in the chord charts of the consumption-side (Figure 8a) and the income-side embodied energy in 2020 (Figure 8b). Figure A1 presents a more detailed version.

Energy flow is measured both on the consumption side and the income side. From the perspective of the consumption side, Hebei, Shanxi, Inner Mongolia, and Liaoning were the top embodied energy export provinces. At the national level, the embodied energy flow of inter-provincial trade was dominated by the Surrounding Bohai Economic Belt and the Yangtze River Economic Belt. The embodied energy flows within the economic belts were unidirectional and asymmetric. In the economic belt, neighboring provinces have close economic and trade relationships, so the scale of embodied energy flow is huge. From the income side, Shanxi, Inner Mongolia, Guangdong, Shandong, Henan, and Jiangsu were the top embodied energy export provinces. From the perspective of the income side, the primary input (raw materials, labor, and capital) of outflow provinces drives the embodied energy consumption of inflow provinces. Shanxi and Inner Mongolia have rich coal resources and are important energy bases of China. Shandong and Henan have dense populations and abundant labor resources. Guangdong and Jiangsu own developed economies and abundant capital. Therefore, from the income side, embodied energy flow exists regional differences because of different location advantages.

4.2. Embodied Energy Flow in Critical Provinces

4.2.1. Inter-Provincial Consumption-Side Embodied Energy Flow

Figure 9 shows the embodied energy driven by each province’s final use of Shanxi. Figure 10 shows the embodied energy promoted by the primary input of Shanxi Province to other provinces. In 2020, the total embodied energy driven by each province’s final use of Shanxi was 80.17 million tons of standard coal, among which the embodied energy driven by Shanxi’s local consumption was 57.91 million tons of standard coal, accounting for 72.2%. It can be seen in Figure 7 that the embodied energy driven by the consumption of Hebei, Inner Mongolia, and Jiangsu to Shanxi was the largest, which is 8.79 million tons of standard coal in all. In 2020, the embodied energy promoted by Shanxi’s primary input was 170.62 million tons of standard coal. Among these, the embodied energy promoted by Shanxi’s primary input was 117.33 million tons of standard coal, accounting for 68.8%. It can be seen from Figure 8 that Hebei, Jiangsu, and Zhejiang have the largest embodied energy promoted by Shanxi’s primary input, which is 29.25 million tons of standard coal in all.

Generally, Shanxi is a net-export embodied energy province. Comparing data in 2015 and 2020, net exports of embodied energy in Shanxi increased from 25.76 million tons of standard coal to 42.55 million tons of standard coal, with an increase in 61%, ranking from fifth to fourth in China. The embodied energy in Shanxi Province shows a trend of flowing to surrounding provinces and economically developed areas. Hebei and Inner Mongolia are adjacent to Shanxi, so the inter-provincial trade is frequent. The consumption-side embodied energy driven by these provinces to Shanxi and the income-side embodied energy attributed to Shanxi from these provinces are huge, and the inter-provincial embodied flow is on a large scale. In addition, Jiangsu and Zhejiang have developed economies and a large scale of production, and the demand for production factors is large. These provinces drive a large amount of consumption-side embodied energy for Shanxi. At the same time, these provinces have developed industries and abundant products. Shanxi consumes products from these provinces, generating a large amount of income-based embodied energy.

4.2.2. Inter-Sectoral Consumption-Based Embodied Energy Flow

As mentioned above, Shanxi is a net embodied energy export province. Figure 11 shows the embodied energy flows of 27 industrial sectors in Shanxi in 2020. The left S1–S27 represents the 27 industrial sectors of Shanxi with embodied energy outflowing to 30 provinces and regions, while the right represents each province, and the thickness of the lines reflects the amount of embodied energy flows. Regarding different sectors, among the top 10 consumption-side embodied energy sectors, there were three energy sectors, PP sector (S11), PE sector (S22), and CM sector (S2); three high energy-intensity sectors, SM sector (S14), CH sector (S12), and NM sector (S13); and the other four sectors are FF sector (S1), TS sector (S25), other sector (S27), and WR sector (S26). Shanxi is an important energy base in China and has a good endowment of energy resources. At the same time, the industrial structure of Shanxi is dominated by heavy industries. According to the CSMAR Database (https://data.csmar.com/ (accessed on 20 September 2025)), in 2020, from the perspective of taxation, the proportion of the second industry’s tax revenue in total tax revenue in Shanxi Province was 62.3%, which was much higher than the national average of 41.8%. At the sectoral level, the tax revenue of the CM sector (S2) in Shanxi Province was 100.5 billion yuan, accounting for 34.5% of the national total; the tax revenue of the SM sector (S14) was 11.6 billion yuan, accounting for 29.0% of the national total; the tax revenue of the PE sector (S22) was 10.2 billion yuan, accounting for 29.3% of the national total; the tax revenue of the PP (S11) sector was 6.3 billion yuan, accounting for 10.4% of the national total. From the employment perspective, the number of employees in the mining industry in Shanxi Province, including CM sector (S2) and MM sector (S4), was about 820,000, accounting for 18.7% of the total employment, which was much higher than the national average of 2.1%; the number of employees in the production and PE sector (S22) in Shanxi Province was about 160,000, accounting for 3.5% of the total employment, which was also higher than the national average of 2.2%. Whether from a taxation perspective or an employment perspective, energy-intensive sectors are an important part of Shanxi Province’s economy, and the production of these high-energy sectors requires a lot of energy. Therefore, in order to meet the production needs of other provinces, Shanxi consumes a large amount of energy to produce products, resulting in a large amount of consumption-side embodied energy in various provinces and regions.

Regarding the consumption-side embodied energy of every province, sectors played different roles. The embodied energy of the SM sector (S14) in Shanxi mainly flowed to Zhejiang, Henan, and Guangdong; the embodied energy of the PP sector (S11) mainly flowed to Henan, Guangdong, and Jiangsu; the embodied energy of PE sector (S22) mainly flowed to Henan, Guangdong, and Hebei; the embodied energy of CM sector (S2) energy mainly flowed to the Hebei, Zhejiang, and Jiangsu. To sum it up, most provinces where the embodied energy from metallurgy, coking, electric power, coal, and other sectors of the Shanxi Province mainly flowed to were China’s important heavy industry bases or manufacturing provinces.

4.3. Embodied Coal Flow

The above analysis of inter-provincial carbon emission flows reveals the complex network characteristics of embodied carbon transfer between regions. It is worth noting that the underlying logic of carbon transfer is strongly correlated with the spatial allocation of energy resources. We merge six kinds of by-products of raw coal and coal, washed coal, other types of coal washing, coal, coke, coke oven gas, and other coking products, and then measure embodied coal according to Equations (9) and (10). By comparing the embodied coal and embodied energy in each province, we find that the proportion of embodied coal and embodied energy on both the consumption and income side of Shanxi Province was more than 50%, ranking first in China. Coal occupies a large proportion of the energy structure of Shanxi, so it is particularly important to study the embodied coal in Shanxi, which provides a typical sample for understanding the spatial transformation of carbon emissions in China. This paper will study the consumption-side and income-side embodied coal of Shanxi in 2015 and 2020, and analyze the consumption and flow of embodied coal in Shanxi.

4.3.1. Total Embodied Coal Consumption

From the perspective of the consumption side, the embodied coal of Shanxi in 2015 reached 37.71 million tons of standard coal, accounting for 52% of its embodied energy, which is higher than the average of 42%. In 2020, the embodied coal in Shanxi reached 34.2 million tons of standard coal, accounting for 50% of the embodied energy, higher than the national average of 38%.

From the perspective of the income side, the embodied coal of Shanxi reached 66.31 million tons of standard coal in 2015, accounting for 61% of the embodied energy, higher than the national average of 53%. In 2020, the embodied coal in Shanxi reached 83.01 million tons of standard coal, accounting for 68% of the embodied energy, higher than the national average of 37%. From 2015 to 2020, the embodied coal and embodied energy of Shanxi Province increased by 25% and 57%, respectively, showing that the energy structure of Shanxi Province was adjusted. Compared with the consumption-side, the embodied coal on the income-side accounted for a larger proportion of embodied energy in Shanxi, indicating that more attention should be paid to the income-side embodied coal in Shanxi.

4.3.2. Coal Flow Embodied in Inter-Provincial Trade

Figure 12a,b, respectively, show the embodied coal flows in Shanxi in 2015 and 2020. Figure 10 has shown that Shanxi is not only an embodied energy export province but also an embodied coal export province. Compared with the flow of embodied energy, the flow of embodied coal in Shanxi was more prominent among the 30 provinces in China. In 2015, Shanxi was the third-largest province with net embodied coal exports. And in 2020, Shanxi became the second-largest province with net embodied coal exports. From the perspective of changing trends, from 2015 to 2020, under the increasing trend of consumption-side embodied energy exports and local consumption of Shanxi, the consumption-side embodied coal and local consumption of Shanxi decreased slightly. The embodied energy imports and embodied coal decreased synchronously. Net embodied coal exports increased by 22% and net embodied energy exports increased by 65%. During the same period, the proportion of Shanxi’s net embodied coal exports to local consumption increased from 79% to 99%, indicating that Shanxi’s provision of products and services to other provinces in inter-provincial trade has an increasing impact on its own coal consumption.

Coal flows can also be divided into the two categories mentioned above: those driven by geographical proximity and those driven by cross-regional industrial demands. Take the 2015 income-based coal of Shanxi, for example, about 2.4 million tons standard coal of the coal flowed to neighboring provinces (Hebei, Henan, Shaanxi, and Inner Mongolia), and about 9.8 million tons of standard coal flowed to developed provinces (Shanghai, Jiangsu, Zhejiang, and Shandong). Both of these two energy flow patterns have significant impacts on the coal consumption in Shanxi.

From the direction of consumption-side embodied coal flow, in 2015, the total embodied coal driven by the consumption of Hebei and Jiangsu to Shanxi was 3.25 million tons of standard coal, accounting for 9%. In 2020, Hebei and Inner Mongolia’s consumption of Shanxi drove embodied coal, adding up to 3.22 million tons of standard coal, accounting for 9%. In general, the main direction of embodied coal flow driven by provincial consumption is from Shanxi to the surrounding provinces.

Figure 13 shows embodied coal driven by every province’s consumption in the 27 sectors of Shanxi. From the perspective of the source sectors of embodied coal in each province, the flow of consumption-side embodied coal mainly occurs in three sectors: the SM sector (S14), PP sector (S11), and PE sector (S22). This is because the energy sectors and the high energy-intensity sectors’ production process demand more coal resources, and also more coal to satisfy inter-provincial consumption.

In terms of consumption-side embodied coal of different provinces, Zhejiang, Henan, Guangdong, Hebei, and Jiangsu drive more consumption-side embodied coal to Shanxi. The source sectors of consumption-side embodied coal of these provinces are also concentrated in three sectors: SM sector (S14), PP sector (S11), and PE sector (S22). It can be seen that the industrial flow of consumption-side embodied coal is consistent with the industrial flow of consumption-side embodied energy.

4.4. Synthesis of Consumption-Based and Income-Based Accounting

Figure 14 presents the comprehensive comparison of embodied energy consumption based on both consumption-side and income-side perspectives in 2015 and 2020. The size of the points is determined by the values of $income-based energy/consumption-based energy$. The larger the value, the larger the point. That is to say, the larger point indicates the greater income-based energy and the smaller consumption-based energy for the corresponding province, and the more responsibility it bears for providing energy to other provinces. The horizontal line in the figure represents the national average income-based energy, and the vertical line represents the national average consumption-based energy. Hebei, Shanxi, and Inner Mongolia were important energy providers, generating a large amount of energy consumption when meeting the production demands of other provinces. Guangdong, Zhejiang, and Jiangsu were important energy consumers, driving the energy consumption of a large number of upstream provinces through consumption. By combining the two perspectives, it can be concluded that economically developed provinces transfer the pressure of energy consumption through the supply chain to energy-rich provinces. Figure 15 shows the comprehensive comparison of embodied coal consumption based on both consumption-side and income-side perspectives in 2015 and 2020. Hebei, Shanxi, and Inner Mongolia were important coal providers, generating a large amount of coal consumption when meeting the production needs of other provinces. Shandong, Guangdong, and Jiangsu were important energy consumers, driving the coal consumption of a large number of upstream provinces through consumption. By combining the two perspectives, it once again proves that economically developed provinces transfer the pressure of coal consumption through the supply chain to coal-rich provinces. Specifically, for Shanxi Province, both income-based energy and income-based coal were significantly higher than the national average, while the consumption-based energy and consumption-based coal were close to the national average. This indicates that Shanxi played an important role in the national energy supply and coal supply, and also assumed more energy responsibilities for the production of other provinces.

4.5. Counterfactual Analysis

Counterfactual analysis was performed to eliminate the influence of COVID-19. Outbreaking in 2020, the COVID-19 pandemic and the subsequent restrictions reduced economic activities, changed production relations, and thus also affected the use of energy and coal. Thus, it is important to apply a counterfactual analysis. According to existing studies [55], the changes in the economic total value and the production changes in different sectors caused by the pandemic were incorporated into the model. In this way, the input–output relationship in 2020 was simulated despite the influence of the epidemic, and the consumption-based and income-based provincial energy consumption and coal consumption were measured. Figure 16 illustrates the comparison of the business-as-usual (BAU) scenario and the counterfactual scenario of energy consumption, and Figure 17 shows the one of coal consumption. Whether energy consumption or coal consumption, the pandemic did not have a significant impact on the consumption patterns of each province. Shandong, Guangdong, and Zhejiang were the main consumers of consumption-based energy. Hebei, Shandong, and Jiangsu were the main energy consumers based on investment. Hebei, Henan, and Jiangsu were the main income-based consumers. And Hebei, Inner Mongolia, and Shanxi were the main consumers of income-based coal. Shanxi played a more important role in inter-provincial coal flow than in inter-provincial energy flow. The main conclusions of this study remain valid. In addition, by comparing the two scenarios, it can be observed that although the output of sectors and the total economic value have increased in the counterfactual scenario, there has been less embodied energy and embodied coal consumption. Therefore, it can be roughly stated that during the pandemic, China’s economic production structure deteriorated, and it became more dependent on the use of coal and energy. The specific measurement of China’s economic structure changes can be further explored in subsequent studies.

5. Conclusions and Suggestions

5.1. Conclusions

This study first estimates the embodied energy of 30 provinces in China in 2015 and 2020 from both the consumption side and the income side. Then, the main provinces and main sectors of embodied energy in Shanxi are analyzed. Last but not least, based on the sufficient coal resources in Shanxi, this paper also analyzes the embodied coal of Shanxi. Shanxi is a critical energy-consumption province, from which both energy and coal flow to economically developed provinces and neighboring provinces. As a net embodied energy and coal exporter, Shanxi bears more energy conservation burden while meeting the demands of other provinces, exemplifying energy inequity and underscoring the need for energy compensation mechanisms in the green transition process.

For the embodied energy among provinces in China, there is a characteristic of uneven spatial distribution. Shandong, Guangdong, Henan, and other provinces with intensive industry and large populations have more embodied energy consumption driven by final use. Of the five elements of final use, fixed capital formation contributed the most to embodied energy consumption. From the perspective of the primary input, Hebei, Guangdong, Liaoning, and other industrial-intensive and resource-abundant provinces have more embodied energy consumption promoted by the primary input. In the four elements of primary input, embodied energy consumption is mainly determined by the compensation of employees and operating surplus. The energy embodied in provincial trade presented a trend so that, from regions that have rich energy and developed heavy industry to regions that lack energy but have developed economies, the flow of embodied energy presents an asymmetry.

For the embodied energy of Shanxi and the whole country, Shanxi is a net embodied-energy export province. From the perspective of provinces, the consumption of neighboring provinces such as Inner Mongolia and Hebei, as well as economically developed provinces such as Jiangsu, drives the most embodied energy to Shanxi. From the perspective of sectors, the consumption-side embodied energy of each province mainly flows from the energy sectors and high energy-intensity sectors, such as the PP sector and MM sector.

For embodied coal in Shanxi and the whole country, the ratios of embodied coal and embodied energy from both the consumption-side and income-side in Shanxi are over 50%, ranking first in China. However, from 2015 to 2020, the ratio declined. Compared with the final use, the primary input contributes more to the embodied coal in Shanxi. In addition, the provincial and sectoral characteristics of consumption-side embodied coal flows driven by provincial consumption in Shanxi are consistent with the characteristics of embodied energy flows. From the provinces, the economically developed provinces such as Zhejiang, Jiangsu, Guangdong, and surrounding provinces such as Hebei and Henan drove the most embodied energy from Shanxi. This embodied energy mainly flows from the SM sector (S14), PP sector (S11), and PE sector (S22) in Shanxi.

5.2. Policy Suggestions

To promote the energy revolution process of Shanxi and to optimize China’s energy consumption structure, based on the results of this study, we mainly put forward the following suggestions.

Firstly, given the asymmetry of energy embodied in the inter-provincial trade, China should care about the coordination and complementarity of regional trade and actively promote the cross-regional “energy compensation” mechanism. At present, provincial sectors pay for the use of energy, while they have not paid for the indirect utility of embodied energy and its environmental costs. So far, many studies have focused on ecological compensation mechanisms, such as carbon compensation [56]. Based on the existing research on carbon compensation, the energy compensation standard can be calculated as follows: $\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=\mathrm{n}\mathrm{e}\mathrm{t} \mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{e}\mathrm{x}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{s} \ast \mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{p}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e} \ast \mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{c}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}$. This study provides a reference for calculating the net energy exports. The energy price is based on the market price. The energy compensation coefficient should take into account the differences in economic development levels among different regions and adjust the compensation standard according to the payment capacity [57]. Energy compensation payers should also be allowed to provide technical or strategic support to the recipients to help improve carbon efficiency and decouple local economic development from carbon emissions [58]. Drawing lessons from these studies, China can implement a cross-regional energy compensation mechanism to provide energy compensation to the embodied energy export provinces presented by Shanxi.

Secondly, from both the consumption and income sides, the surrounding provinces are major areas where embodied energy and coal in Shanxi flow. Therefore, the Shanxi provincial government should actively promote coordinated development with surrounding provinces and apply for the construction of a national demonstration zone for undertaking industrial transfer. The economically developed provinces are also important outlets for energy and coal consumption in Shanxi. Coastal regions, through means such as technology transfer and industrial cooperation, assist Shanxi in enhancing resource utilization efficiency and industrial added value, promoting its green and low-carbon transformation. Through these ways, Shanxi can promote energy efficiency and accelerate the low-carbon energy transition of a resource-based economy.

Thirdly, since the inter-provincial flow of embodied coal in Shanxi mainly comes from energy-intensive sectors and high-energy-consumption sectors, the provincial government should guide these energy-related sectors, such as metallurgy, coking, electric power, and coal mining, to carry out energy transformation [59]. The government should build a great energy data platform, upgrade the energy system, and optimize the structure of the energy industry to provide scientific support for energy development in Shanxi Province.

5.3. Shortcomings

There are several shortcomings in this study. Firstly, exports should also be an important indicator when studying embodied energy from the final use perspective. However, since the export item of the input–output table used in this paper lacks more detailed content, this paper only considers the domestic inter-provincial trade. Due to the lack of export factors, consumption-based embodied energy and coal may cause errors. Secondly, this paper only analyzes the total consumption and flow of embodied energy and does not further analyze the influencing factors of embodied energy. In later studies, it can be incorporated into study content to enrich the research on embodied energy. Thirdly, sector aggregation into 27 categories may introduce uncertainty in structural analysis [60]. More detailed trade data at the sectoral level could improve the accuracy of regional energy and coal calculations. The missing data of Tibet’s energy and coal consumption will not have a significant impact on the result, but if there are updated data available for supplementation, it will improve the measurement accuracy. Fourthly, the RAS method also caused some bias when calculating economic relationships (See Appendix E), even though it is the most commonly used input–output table-updating method. It assumes that the changes in the economic structure are smooth and gradual, so it cannot cover technological coefficient changes. If there are updated and more detailed data in the future, the research can be further expanded. For the deficiencies mentioned above, a more comprehensive study could be conducted in the future if updated and more detailed data are available.