Measurement of carbon emissions and responsibility sharing for the industrial sector in Zhejiang, China

Abstract

Carbon reduction is imperative for achieving carbon peaking and neutrality. Accordingly, it is important to determine which industrial sectors have more responsibility in this matter. Based on data from Zhejiang's input-output tables, this study applies the Input-Output method to measure and compare the carbon emissions of 42 industrial sectors in Zhejiang Province from 2002 to 2017, and then assesses the carbon emission attributes of each industrial sector, and ultimately determines the responsibility for carbon emission reduction from the perspectives of the producers and consumers. The results of the study show that direct carbon emissions and whole-process carbon emissions in Zhejiang increased continuously from 2002 to 2017, with carbon emission intensity first decreasing and then increasing. However, carbon emission intensity was much lower in 2017 than in 2002. Over time, the attributes of carbon emissions by sector changed little. Particularly, high-carbon sectors covered most of the energy supply sectors, low-carbon sectors were mostly tertiary-related, and pseudo-low-carbon sectors were mainly found in the productive services sector. In terms of carbon emission reduction responsibilities, there are large differences in emission reduction responsibilities between sectors, with the electricity and heat sectors bearing the largest responsibilities based on their production, consumption and total carbon emission reductions. The conclusions of this study can provide some data support for the further development of carbon peak and carbon neutral plans.

1. Introduction

According to an analysis by the International Energy Agency in 2021, global CO₂ emissions from energy consumption reached 36.3 billion tonnes, with China emitting over 11.9 billion tonnes, representing 33% of the total. China has a large economy and thus, a large demand for energy. However, energy usage plays a pivotal role in global climate change and urgent changes are required to reduce the CO₂ emissions associated with specific energy sources. The report of the 20th National Congress of the Communist Party of China highlighted the need to actively and steadily promote carbon peaking and neutrality, which is a broad and profound economic and social change [1]. Accordingly, Zhejiang has set out the ‘CO₂ emission peak action’ column from the 14th Five-Year Plan, proposing measures from six aspects, including the adjustment of the energy structure and promotion of low-carbon industries. The share of non-fossil energy in primary energy consumption in Zhejiang during the 13th Five-Year Plan period was 20.3%, and the share of the tertiary sector in the gross domestic product (GDP) was 55.8%, indicating that the economic development of Zhejiang is gradually decarbonising. Still, most available data are at the macro level, without accounting for the specific situation in each industrial sector. Therefore, this study aimed to measure and compare the carbon emissions of industrial sectors in Zhejiang from 2002 to 2017, assess the carbon emission attributes of each industrial sector, and determine the carbon emission reduction responsibilities from the perspectives of producers and consumers. The results of this study provide a reference for policy- and decision-makers in working towards achieving a green and low-carbon modern industrial pattern in Zhejiang.

Associations among carbon emissions, economic growth, and industrial structure have been studied by many domestic and international scholars: Proops [2] constructed models from 10 sectors excerpted from input-output tables to study the impact of final demand on economic growth, energy consumption, and environmental pollution, and the relationship between them. Miller and Blair [3] used the same input-output approach to analyse energy consumption and environmental pollution and explored the impact of and relationships among economic development, energy consumption, and environmental pollution. Zhu et al. [4] demonstrated a U-shaped relationship between energy mix transition and economic growth, and an inverted U-shaped relationship between CO₂ emissions and economic growth. Yuan and Sun [5] explored the complex relationship between industrial structure, energy consumption, economic growth, and carbon emissions based on the level of economic development of each Chinese province. Moreover, many scholars have focused on the factors influencing carbon emissions: Weber and Perrels [6] used methods such as scenario analysis to examine the impact of lifestyle changes on energy demand, and hence on changes in carbon emissions. Hastuti et al. [7] used an input-output analysis framework to analyse the main drivers of change in CO₂ emissions in Indonesia. Using data on per capita carbon emissions in Chinese counties, Wang et al. [8] found that the secondary industry output volume and carbon intensity are significantly and positively correlated with carbon emissions, while population density and government expenditure are significantly and negatively correlated. Furthermore, many researchers have investigated carbon emission measurement methods and research levels, which include the Intergovernmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas Inventories scenario analysis [9], energy consumption per unit of GDP [10], life cycle analysis [11], and input-output analysis [12,13]. Life cycle and input-output analyses can penetrate the meso-industrial sector and micro-firm levels. Research on carbon emissions mainly includes the single country [14], Inter-provincial [15,16], specific industry levels，such as Construction industry [17]，Transport sector [18]，Coal industry [19] as well as textile and apparel industry [20], and international trade [21,22].

As for the division of carbon emission responsibility, a variety of approaches have been explored. According to the prevailing ‘polluter pays’ principle, ‘producer responsibility’ can be deduced. This is also in line with the ‘territorial responsibility’ principle recognised by international organisations such as the IPCC and international treaties such as the Kyoto Protocol and the Paris Agreement. Furthermore, based on the principle of ‘the one who benefits should provide compensation’, consumers should also be held responsible for carbon emissions. Hence, this raises the question of how much responsibility consumers should bear. The consumer responsibility theory addresses this, as carbon footprint accounting based on consumption laws has progressively become the basis for policy formulation [[23], [24], [25], [26]]. Theoretically, a more reasonable model would be the principle of ‘shared responsibility’, in which Ferng [23] proposed a weighted calculation based on the original ‘consumption-benefit’ and ‘production-benefit’ principles. However, this approach has faced critique over issues with double counting. Bastianoni et al. [24] proposed the carbon emission augmentation (CEA) method as a new approach for responsibility allocation; yet it lacks practical applications. Lenzen et al. [27] determined weights based on the value share added by each segment to the external inputs of industry and then measured the carbon responsibility of producers and consumers.

Through the existing literature, it can be found that few studies have examined carbon emissions from all sectors in a particular region, and in terms of carbon emission responsibility identification, studies on the distribution of responsibility for carbon emissions in specific industrial sectors are lacking [28]. Past studies are primarily based on the ‘producer’ perspective and have not looked at the whole industry chain, which can lead to problems such as ‘carbon leakage’. Across the chain, any sector may simultaneously receive carbon emissions from the upstream sector and transfer them to the downstream sector. In actual production and consumption, carbon emissions from the upstream sector are not completely transferred to the downstream sector. Therefore, this study (1) took all the sectors in Zhejiang Province as the object of study and measures carbon emissions from individual sectors as well as carbon emissions flowing between sectors. (2) defined the carbon emissions that remain within the sector as the responsibility of the production chain, and the carbon emissions that are transferred to the downstream sector as the responsibility of the consumption chain. Based on this, we divided the responsibility for carbon emissions among the various sectors in Zhejiang.

2. Methods

2.1. Study aim

This study applied the input-output method, which uses a combination of input-output table and energy use data from Zhejiang from 2002, 2007, 2012, and 2017. Our aims are as follows: (1) To measure the direct and whole-process carbon emissions of 42 industrial sectors in Zhejiang from 2002 to 2017 and make horizontal and vertical comparisons. (2) Define the carbon emission attributes of each industrial sector by measuring carbon emission coefficients and clarify the key sectors for reducing emissions. (3) Confirm the responsibility for carbon emissions in terms of production and consumption of each industrial sector and clarify the responsible links for reducing emissions. The research framework is shown in Fig. 1, and the specific details of the research methods are shown below.

Fig. 1.

Research flow chart.

2.2. Calculation of carbon emissions

In order to measure the direct and indirect carbon emissions of various industries in Zhejiang Province, and to clarify the structure of carbon emissions as well as the sharing of responsibility for carbon emission reduction, we apply the input-output method on the basis of the IPCC method.We mainly used the input-output tables of Zhejiang from 2002 to 2017, China Energy Statistics Yearbook, Zhejiang Statistics Yearbook, and relevant data from the National Bureau of Statistics to measure carbon dioxide emissions from fossil fuel combustion, as described in the following subsections.

2.2.1. Estimation of apparent energy consumption

The energy balance does not provide detailed micro data by industry. Thus, we used the segmentation data and energy prices of each industrial sector in the input-output tables to estimate their energy consumption. Then, we added up the consumption of various energy sources to obtain the total consumption of a sector. Moreover, as the consumption was expressed in a variety of original units, it was necessary to convert these (tonnes, cubic meters, etc.) into a common unit of energy (TJ). The calculation formulas are as follows:

$${AC}_j^{\prime }=\sum _{k=1}^1{AC}_{jk}^{\prime }=\sum _{k=1}^1{AC}_{jk}\times F_k=\sum _{k=1}^1\frac{V_j\times A_{ij}\times r_{ik}}{P_k}\times F_k$$ (1)

$${BC}_j^{\prime }=\sum _{\mathrm{k}=1}^1{BC}_{jk}^{\prime }=\sum _{\mathrm{k}=1}^1{BC}_{j\mathrm{k}}\times F_{\mathrm{k}}=\sum _{\mathrm{k}=1}^1\frac{V_j\times A_{\mathrm{i}j}\times r_{\text{ik}}}{P_{\mathrm{k}}}\times F_{\mathrm{k}}$$ (2)

Here, AC_jk is the apparent consumption of energy k directly consumed by department j, BC_jk is the apparent consumption of energy k consumed by the entire process in department j, k = 1, 2, …, l, V_j is the output value of sector j, A_ij is the direct consumption coefficient of sector j to energy sector i (i = 1, 2, …, n; j = 1, 2, …, m), B_ij is the complete consumption coefficient of sector j to energy sector i (i = 1, 2, …, n; j = 1, 2, …, m), r_ik is the proportion of energy k output value in energy sector i, P_k is the price of energy k, and F is the conversion factor expressed as the average low heating level of energy k.

2.2.2. Calculation of the net carbon content of each sector

The net carbon content is the total carbon content minus the non-energy-use portion of fossil fuels that produce greenhouse gases without combustion. The formulas are as follows:

$$NAC=\left(1-{NEU}_k\times f_k\right)\times {AC}_j^{\prime }\times c_k$$ (3)

$$NBC=\left(1-{NEU}_k\times f_k\right)\times {BC}_j^{\prime }\times c_k$$ (4)

Here, NAC is the net carbon content of direct consumption in each sector, NBC is the net carbon content of whole-process consumption in each sector, c_k is the average carbon content of energy k (tonnes C/TJ), NEU_k is the proportion of non-energy utilisation, and f_k is the carbon sequestration rate of the non-energy utilisation sector; the remaining variables are the same as in previous formulas.

2.2.3. Calculation of the carbon emissions of each department

In real life, the carbon content of the apparent energy consumption is not fully oxidised to produce CO₂ emissions. Therefore, the oxidation process must be corrected to convert carbon emissions to CO₂ emissions. The formulas are as follows:

$${CF}_a=\sum _{k=1}^1{NAC}_k\times {OF}_k\times 44∕12\text{,}$$ (5)

$${CF}_b=\sum _{k=1}^1{NBC}_k\times {OF}_k\times 44∕12$$ (6)

Here, CF_a is the direct carbon emission, CF_b is the whole-process carbon emission, OF_k is the carbon oxidation coefficient, and 44/12 is the conversion coefficient to CO₂ emissions; the remaining variables are the same as in previous formulas.

2.3. Calculations for the determination of carbon emission characteristics

The energy consumption and carbon emissions of each industrial sector in Zhejiang were calculated as shown above. However, these are aggregates and do not visually reflect the high- or low-carbon attributes of a particular industry. Therefore, we introduced the concept of a carbon emission factor. The carbon emission factor is the amount of carbon emitted per unit of the total output in sector i. The formula is as follows:

$$m_i^L={CF}_i^a/X_i$$ (7)

Here, CFαi is the total direct carbon emissions of sector i (i = 1, 2 …, S) and X_i is the total output of sector i.

These factors are defined in terms of direct carbon emissions and are referred to as direct carbon emission factors. However, there is also a group of industries that do not consume much energy or emit much carbon, but whose upstream industries are high-energy-consuming and high-emission. These industries contain more ‘embodied energy’ and ‘embodied carbon’ than direct carbon. Therefore, a whole-carbon emissions factor must be introduced. The formula is as follows:

$$m_i^{\alpha }=\sum _j{d}_j{b}_{ji}$$ (8)

Here, d_j is the element in the direct carbon emission coefficient vector, b_ji is the consumption of sector i's total output by the production of a unit of the final product in sector j, (I-A)⁻¹ (I represents the unit matrix) is the Leontief inverse matrix, A is the direct consumption coefficient matrix, M is the complete consumption coefficient matrix, (I-A)⁻¹ = M + I, m^α = m^L (I-A)⁻¹ = (mα1, mα2 ∼ mα 42) is the full emission factor vector, and m^L = (mL 1, mL 2 ∼ mL 42) is the direct emission factor vector.

2.4. Calculation of the share of carbon emission responsibility

In this study, we set the share of carbon emission reduction responsibility of the consumption links as α_i, whose value is equal to the proportion of complete carbon emissions transferred out of sector i to the total complete carbon emissions of sector I; 1-α_i is the share of carbon emission reduction responsibility in the production link, X_i-x_ij are net inputs (total inputs minus self-supplied inputs), and x_oi is the value added of the industrial sector. The formula is as follows:

$$1-{\alpha }_i=\frac{x_{oi}}{X_i-{\mathrm{x}}_{\text{ij}}}，\mathrm{i}=1\sim \mathrm{S},\mathrm{j}=1\sim \mathrm{S}$$ (9)

The carbon reduction responsibility of industrial sector i for the consumption link (${\mathit{U}}_{\mathit{i}}^{\mathit{\alpha }\mathit{c}}$) is equal to the whole-carbon emissions transferred from that sector to the final demand sector. The carbon reduction responsibility of the production link (${\mathit{U}}_{\mathit{i}}^{\mathit{\alpha }\mathit{p}}$) is equal to the whole-carbon emissions of sector i trapped in the carbon emissions of that sector. The formulas are as follows:

$$U_i^{\alpha c}={\alpha }_i{m}_i^{\alpha }{x}_{oi},i=1\sim S$$ (10)

$$U_i^{\alpha p}=\left(1-{\alpha }_i\right)m_i^{\alpha }{X}_i,i=1\sim S$$ (11)

Here, α_i is the share of carbon reduction responsibility of the consumption link and mαi is the whole-carbon emission factor for sector i. Then, the carbon reduction responsibilities of the production and consumption links were added to form the total carbon reduction responsibility of the industrial sector. The formula is as follows:

$${\mathit{U}}_{\mathit{i}}^{\mathit{\alpha }}={\mathit{U}}_{\mathit{i}}^{\mathit{\alpha }\mathit{p}}+{\mathit{U}}_{\mathit{i}}^{\mathit{\alpha }\mathit{c}},\mathit{i}=\mathbf{1}\sim \mathit{S}$$ (12)

3. Results

3.1. Calculation of carbon emissions

3.1.1. From the perspective of total carbon emissions

Fig. 2 shows the trends in total carbon emissions and carbon emission intensity (carbon emissions per unit of GDP) in Zhejiang from 2002 to 2017. Total carbon emissions from direct and whole-process energy consumption based on the total output of 42 sectors in Zhejiang increased continuously, with an average growth rate of 8.11% for direct energy consumption CO₂ emissions and 12.08% for whole-process energy consumption CO₂ emissions from 2002 to 2017. This increase may have been related to the high rate of economic development in Zhejiang during the study period, when the province's GDP grew at an average annual rate of 13.45%. During periods of economic expansion, growth-related activity inevitably leads to the release of large amounts of CO₂, especially under the influence of factors such as poor industrial structure and low energy efficiency. The increase in carbon emissions in Zhejiang in 2017 was in line with the nationwide trend in carbon emissions, which began to slow down in 2016 and then regained growth in total emissions in 2017. Such variability leads to uncertainty regarding future carbon emissions [29]. Carbon emission intensity can, to some extent, reflect the logical structure of the economy and the level of technology. Fig. 2 also shows that Zhejiang's carbon intensity fluctuated, but with an overall downward trend. The direct carbon emissions intensity in 2017 decreased by 50.15% compared to that in 2002. Both direct and whole-process carbon emission intensity showed a decreasing trend followed by an increasing trend, which corresponds with the historical pattern of economic development in Zhejiang. At the beginning of the Reform and Opening-up, Zhejiang had been vigorously developing light industry, considering the characteristics of the province's resource endowment. Since the beginning of the 21st century, the Internet and financial (FIN) industry has been developing rapidly, the proportion of service industries has been increasing, and carbon emission intensity has been decreasing. In 2016, the government of Zhejiang released the ‘13th Five-Year Plan for Manufacturing Development’, which focused on developing the economy and growing the manufacturing sector, leading to an increase in carbon emission intensity. According to Hoffman's theorem [30], the ratio of net output in the consumer goods sector to that in the capital goods sector must gradually decline; therefore, concerned about the current rebound in carbon emission intensity is unnecessary, as it is bound to decline as industrialisation progresses.

Fig. 2.

Total carbon emissions and carbon emission intensity in Zhejiang in 2002–2017.

3.1.2. From the perspective of industrial structure

Table 1 shows the direct CO₂ emissions and their share from the three sectors (primary, secondary, and tertiary) in Zhejiang from 2002 to 2017. From a horizontal comparison perspective, the secondary sector accounts for the largest share of direct carbon emissions. From a longitudinal comparison perspective, from 2002 to 2012, the share of direct carbon emissions from the secondary sector trended upward, while the tertiary sector trended downward. However, from 2012 to 2017, the share of direct carbon emissions from the secondary sector declined, while the share of the tertiary sector increased. In 2014, the proportion of tertiary industries exceeded that of secondary industries at a ratio of 4.42:47.73:47.84 (primary:secondary:tertiary sectors). For the first time in Zhejiang's history, the industrial structure was characterised by ‘three, two, one’ (tertiary, secondary, primary), signalling that Zhejiang had entered a service-oriented information society. In 2015, the government of Zhejiang issued the ‘Implementation Opinions on Accelerating the Development of Productive Service Industry to Promote the Adjustment and Upgrading of Industrial Structure’, a document that calls for accelerating the optimisation and upgrading of industrial structure. Since then, economic activities have heightened and carbon emissions have increased, which has led to a change in the share of direct carbon emissions between the secondary and tertiary industries.

Table 1.

CO₂ emissions from direct energy consumption of total output of three sectors in Zhejiang (2002–2017).

Year	Primary sector		Secondary sector		Tertiary sector
	Direct carbon emissions	Percentage	Direct carbon emissions	Percentage	Direct carbon emissions	Percentage
2002	0.07	1.91%	3.21	90.79%	0.26	7.30%
2007	0.04	0.87%	3.88	92.22%	0.29	6.92%
2012	0.01	0.09%	5.91	94.77%	0.32	5.15%
2017	0.02	0.20%	10.50	92.01%	0.89	7.78%

Note: direct carbon emissions are provided as billions of tonnes.

3.1.3. From the perspective of the industrial sector

According to the sectoral rankings of direct carbon emissions shown in Table 2, Electricity and heat production and supply (EH), petroleum (PL) processing, coking and nuclear fuel processing (PCN), and chemical industry (CI) remained the main sources of direct carbon emissions as of 2017. The rankings of the other industrial sectors changed, but not substantially. According to the sector ranking of whole-process carbon emissions, between 2007 and 2017 (except for 2002) the top three whole-process energy consumption carbon emission levels in Zhejiang were in the EH, chemical manufacturing, and construction industries. In 2002, the textile (TL) industry ranked third in terms of whole-process carbon emissions, followed by the chemical industry, and the EH. The highest-ranking sectors in all study years were those related to energy supply.

Table 2.

Direct and whole-process carbon emissions of the top five sectors each year (2002–2017)^a.

Rank	2002		2007		2012		2017
	Department	Direct	Department	Direct	Department	Direct	Department	Direct
1	EH	1.16	EH	1.67	EH	2.80	EH	0.02
2	PCN	0.46	PCN	0.64	PCN	0.74	PCN	5.96
3	CI	0.38	CI	0.37	CI	0.64	CI	1.18
4	NMP	0.22	NMP	0.31	NMP	0.30	TSP	0.59
5	TL	0.19	TSP	0.15	MSR	0.27	TL	0.40
Rank	2002		2007		2012		2017
	Department	Whole- process	Department	Whole- process	Department	Whole- process	Department	Whole- process
1	CI	1.30	EH	2.38	EH	5.80	EH	0.62
2	EH	1.27	CI	1.65	CI	4.28	CI	14.42
3	TL	0.97	AT	1.23	AT	2.67	AT	6.67
4	AT	0.84	TL	0.92	TL	1.87	TL	4.57
5	GSE	0.58	GSE	0.87	MSR	1.80	EME	4.19

Carbon emissions are provided as 100 million tonnes.

^aAbbreviations: EME: Electrical machinery and equipment, GSE: General and special purpose equipment manufacturing, TSP: Transport, storage, and postal.

3.2. Determination of carbon emission characteristics

Table 3 shows the classification of the carbon emission characteristics of the 42 sectors in Zhejiang from 2002 to 2017. In general, industrial sectors can be divided into three categories: One category includes high-carbon sectors with evident and stable carbon emission characteristics, such as the EH, PCN, non-metallic mineral products (NMP), non-metallic and other ore mining products (NMM), chemical manufacturing, and metal smelting and rolling products (MSR) industries. Most of these industrial sectors belong to energy-heavy and manufacturing industries. Thus, they are big carbon emitters. The second category includes low-carbon sectors, such as information transmission, software, and information technology services (ISI); FIN; and scientific research and technical services (SRT), which are in the bottom half of the carbon emission factors. These are typically relatively low-carbon industries in terms of both direct and whole-process emission factors, and they are basically tertiary industries. The third category includes sectors in which the characteristics of carbon emissions depend on the measurement perspective and where direct emissions differ substantially from whole-process emissions. These are sectors with large whole-process carbon emissions that also take on a large amount of carbon emissions from external sectors in the interaction of product chains and are pseudo-low carbon sectors. During the study period, the largest changes from direct to whole-process emission factors were found in the productive services and manufacturing sectors, among others.

Table 3.

Classification of carbon emission characteristics of 42 sectors in Zhejiang (2002–2017)^a.

Year	High-carbon sectors	Low-carbon sectors	Pseudo-low carbon sectors
2002	WS; PL; CNP; MW; MES; EH; NMP; CI; NMM; MSR; MOP; etc.	TSP; ISI; FIN; RE; SRT; WEU; PSS; etc.	WPF; TSP; ISI; LBS; HSW; etc.
2007	WS; PL; CNP; MW; MES; NMP; NMM; EH; CI; PPE; MSR; etc.	FIN; TSP; ISI; EDU; SRT; WR; RRO; etc.	TSP; EME; SE; GP; HSW; WPF; OMP; etc.
2012	CMP; EH; GP; PL; CNP; OMP; NMP; NMM; CI; WPS; MSR; etc.	FIN; RE; SRT; PSS; ISI; EDU; AFF; etc.	EME; IM; CCE; MW; HSW; GE; etc.
2017	EH; CMP; GP; PL; CNP; NMP; MSR; TL; CI; NMM; etc.	PSS; FIN; REV; ISI; WR; AFF; CSR; RE; etc.	WPS; HSW; SE; WEU; MW; MES; EME; etc.

^aAbbreviations: CSR: Culture, sport, and recreation, GE: General equipment, CCE: Communications equipment, computers, and other electronic equipment, IM: Instruments and meters, AFF: Agriculture, forestry, and fishery products and services, WPS: Water production and supply, CMP: Coal mining products, OMP: Other manufacturing products, SE: Specialised equipment, GP: Gas production and supply, RRO: Residential services, repairs, and other services, EDU: Education, PPE: Paper, printing, and educational and sporting goods, HSW: Health and social work, LBS: Leasing and Business Services, WPF: Woodworking products and furniture, PSS: Public administration, social security, and social organisations, WEU: Water, environment, and utilities management, RE: Real estate, MOP: Metal ore mining products, MES: Machinery and equipment repair services, MW: Metalwork, CNP: Coking products and processed nuclear fuel products.

3.3. Share of carbon emission responsibility

Here, we compare the carbon reduction responsibilities of each sector from the perspectives of both production and consumption links. The top 5 sectors in Zhejiang in terms of carbon emission reduction responsibility in the production link from 2002 to 2017 are shown in Table 4. In this period, the largest responsibility for carbon reduction regarding production was in the Waste and scrap (WS), CI, Textiles (TL), GP sectors. Starting in 2012, the EH sector became the largest responsible sector and considerably exceeded the rest in terms of value. However, the carbon reduction responsibility of this sector was 515 million tonnes in 2017, which was 91.22% higher than that in 2012. Moreover, this sector is upstream in the industrial chain and is part of the energy supply sector, undertaking the province's production of electrical energy and transferring large amounts of CO₂ while delivering electricity and heat to downstream industries. Since a large proportion of this CO₂ is trapped within this sector, it needs to take more responsibility for carbon reduction in the production link. In addition to the EH sector, the CL, architecture (AT), and TL sectors were responsible for prominent carbon reductions between 2012 and 2017. In 2017, the responsibility for carbon emission reductions regarding production in service sectors, such as transport, storage, and postal (TSP); leasing and business services (LBS); and wholesale and retail (WR), also increased substantially, which may have been related to the accelerated development of production services in Zhejiang during industrial restructuring.

Table 4.

Responsibilities for carbon emission reduction in production links of all sectors from (2002–2017).

Rank	2002		2007		2012		2017
	Sector	Responsibility for the production link	Sector	Responsibility for the production link	Sector	Responsibility for the production link	Sector	Responsibility for the production link
1	WS	0.7318	WS	1.2007	EH	2.6927	EH	5.1491
2	CI	0.6342	CI	1.0265	CI	1.2013	CI	3.2532
3	TL	0.4620	TL	0.5141	AT	0.7329	TL	1.5957
4	GP	0.2715	GP	0.3639	TL	0.6337	AT	1.3450
5	AFF	0.2576	MSR	0.3574	MSR	0.4781	TSP	1.2462

Responsibilities for carbon emissions production are provided as 100 million tonnes.

In terms of the responsibility for carbon reduction in the consumption link, the WS, CI, gas production and supply (GP), and TL sectors were also the most responsible between 2002 and 2007. From 2007 to 2017, the results were the same as those regarding responsibility in the production link. That is, the EH also became the sector with the greatest responsibility to reduce emissions in the consumption link. This is because this sector is part of the energy supply sector and has a close relationship with the various sectors downstream in the industry chain. All industries need energy products from this sector, and the portion of carbon emissions delivered is the responsibility of the consumption link. Another sector that bears great responsibility to reduce carbon emissions is the AT, which relies on a large amount of energy and resources for the raw materials needed for construction and produces carbon emissions during the use of the buildings themselves. As a result, the AT generates a large amount of implied carbon emissions from the production link to the consumption link, which in turn is transferred to downstream industries. Therefore, this sector bears most of its responsibility for carbon reduction from the consumption link. Simultaneously, in 2017, the carbon reduction responsibilities of the productive service sectors, such as LBS and TSP, increased markedly in terms of consumption.

The total carbon reduction responsibility of each industrial sector was obtained by summing up the carbon reduction responsibility from each production and consumption link. The WS, CI, and TL sectors had the greatest total carbon reduction responsibilities between 2002 and 2007. The EH, CI, and AT were the sectors with the largest total carbon reduction responsibilities between 2012 and 2017. The largest increases in total carbon reduction responsibility were in the EH and AT sectors between 2007 and 2017.

We determined that the above sectors are major consumers of resources and energy. Thus, they are also key areas for energy saving and emission reduction. These industrial sectors are closely linked to other industrial sectors and generate large amounts of embedded carbon emissions in their production and consumption, which are transferred to final consumption as they are produced and consumed between industrial sectors. Table 5 shows the carbon emission reduction responsibilities for the production and consumption links in Zhejiang from 2002 to 2017 and their proportions. During the study period, the production link always had a greater share of responsibility for carbon reduction than the consumption link. Therefore, the production sector should be the key industrial sector implementing carbon reduction measures.

Table 5.

Share and proportion of carbon emission reduction responsibilities (2002–2017).

Year	Responsibility for the production link (billion tonnes)	Proportion	Responsibility for the consumption link (billion tonnes)	Proportion	The total responsibility (billion tonnes)
2002	5.03	67.91%	2.38	32.09%	7.41
2007	6.61	69.22%	2.94	30.78%	9.56
2012	11.27	66.22%	5.75	33.78%	17.02
2017	26.62	69.31%	11.79	30.69%	37.78

4. Discussion

By measuring and analysing the carbon emissions of Zhejiang Province in 2002, 2007, 2012 and 2017, it is found that the total carbon emissions of Zhejiang Province have continued to increase from 2002 to 2017, the intensity of carbon emissions has fluctuated and decreased, and the secondary industry plays a decisive role in the total carbon emissions. On the one hand, since the establishment of the 12th and 13th Five-Year Plans, Zhejiang has experienced rapid economic development, and industrial development has been dominated by light industries such as CI and TL. This type of industry has a high demand for electricity, heat, fuel, and other types of energy, which in turn makes direct carbon emissions higher. On the other hand, whole-process carbon emissions focus on the implicit carbon emissions that exist in the sectors. The top sectors are generally located in the middle of the chain, which means that they are closely related to the other sectors in the industry chain and are prone to take on a large amount of implicit carbon emissions from the upstream sectors, resulting in large whole-process carbon emissions.

As shown in Table 3, the characteristics of carbon emissions by sector changed little between 2002 and 2017, with only a few industrial sectors showing changes in their carbon emission characteristics. The agriculture, forestry, and fishery product and service (AFF) sector were classified as a low-carbon sector in 2012 and 2017, which may be due to government policies. In 2010, the government of Zhejiang issued the ‘Opinions on Further Strengthening Comprehensive Agricultural Development Work’ policy, and in 2014, the ‘Opinions on Accelerating the Development of Modern Ecological and Circular Agriculture’ policy. These policies required a shift in agricultural development mode, advancements in agricultural science and technology, deliberate development and utilisation of agricultural resources, and modernization of agricultural technology. Consequently, agriculture in Zhejiang has shifted from a high-carbon sector to a low-carbon sector.

Overall, whole-process carbon emission factors are a more reasonable measure of the actual carbon emissions of each industrial sector than direct carbon emission factors and are more informative. From the perspective of energy conservation and emission reduction, more attention should be paid to pseudo-low carbon sectors with large whole-process carbon emissions. Strengthening the control of energy consumption and carbon emissions in such industries is key to achieving energy conservation and emission reduction.

5. Policy implications

Based on our analysis, we recommend the following: Firstly, the peak carbon targets for key industrial sectors should be clarified and they should be supported to take the lead in reaching carbon peaking. With eight critical years left to achieve China's committed carbon peaking target, industries with high carbon emissions and intensities, such as EH, AT, TL, and TSP, should be designated key sectors of concern. Specific peak targets should be clearly defined, and more stringent assessment criteria implemented to guide key enterprises in the industry to actively transition to a green, low-carbon development model. Secondly, in the last few years, the structure of the three industries has been optimised, which has accelerated the pace of reducing carbon emissions and intensity. Since 2014, the industrial structure of Zhejiang has been upgraded to ‘three, two, one’, and a pattern of industrial sectors and service industry jointly driving economic development has been formed, but the tertiary sector still has great potential for development. It is important to note that industries such as health and social work (HSW); AFF; FIN; culture, sport, and recreation (CSR); LBS; and education (EDU) are among those that might appear low-carbon but are high-carbon and require consideration in this development process. These industries contain a high level of implied carbon but are important enough that they cannot be restricted in their development, let alone phased out, and can only be developed with enhanced energy efficiency improvements.

6. Limitations

This study has certain limitations, which need to be further explored to improve it. The input-output approach is one of the most common methods for calculating carbon emissions. Compared with the total amount of carbon emissions, this method focuses more on the issue of carbon flows between sectors and the structure of carbon emissions. This paper takes Zhejiang Province as an example to explore the carbon emission structure of various industries. In the future, it can be extended to the whole country to study the carbon emission structure of each industry from the national perspective. The study has identified the carbon reduction responsibilities of various sectors in Zhejiang Province from both the consumption and production links. It is important to explore more deeply the factors influencing carbon emissions and the comprehensive economic benefits of carbon emissions, to propose a more specific scenario for Zhejiang Province to achieve carbon peak and carbon neutrality.

7. Conclusions

Based on the input-output tables and energy consumption data in Zhejiang from 2002 to 2017, this study used the input-output model to measure the direct carbon emissions and the whole-process carbon emissions of 42 industrial sectors in Zhejiang, which were classified into high-carbon, low-carbon, and pseudo-low-carbon sectors according to their carbon emission characteristics. The carbon emission reduction responsibilities of each industrial sector were determined from the production and consumption links of the production chain. The conclusions of the study are as follows.

• (1)From 2002 to 2017, the total carbon emissions in Zhejiang continued to increase, but the carbon emission intensity showed a fluctuating downward trend. The total direct and whole-process carbon emissions increased in these 15 years, with the whole-process carbon emissions increasing more than the direct carbon emissions. In terms of carbon emission intensity, direct whole-process carbon emission intensity fluctuated, but generally showed a downward trend during the study period. In terms of the gap between the three sectors (primary, secondary, and tertiary), the secondary sector plays a decisive role in the total carbon emissions and has great potential for carbon reduction.
• (2)There was little change in the proportion of carbon emissions among the three major carbon emission (high-, low-, and pseudo-low-carbon) sectors. The results showed that the energy, heavy, and manufacturing industries are primarily high-carbon sectors, those in the tertiary sector are mostly low-carbon sectors, and those related to productive services are mostly pseudo-low-carbon sectors. Therefore, the key to energy saving and emissions reduction is to reduce carbon emissions from pseudo-low-carbon sectors.
• (3)Zhejiang had a greater responsibility for carbon emission reduction in the production link than in the consumption link. Further, the responsibility for emissions reduction varies widely by sector. Of the carbon emissions in Zhejiang, the production link is responsible for 70% and the consumption link is responsible for 30%, with no significant change between 2002 and 2017. This indicates that Zhejiang could achieve more effective emission reduction results by starting with the production link and developing a reasonable emission reduction plan.

Funding statement

Not applicable.

Data availability statement

The data can be requested by contacting the authors if required.

CRediT authorship contribution statement

Mengting Shao: Writing – review & editing, Writing – original draft, Resources, Data curation. Xuewang Dong: Supervision, Conceptualization. Huating Huang: Writing – review & editing, Data curation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.