Differential Quantitative Analysis of Carbon Emission Efficiency of Gansu Manufacturing Industry in 2030

Abstract

Decomposition analysis and forecasting of carbon emissions in manufacturing are crucial for setting sustainable carbon-reduction targets. Given the varied carbon-emission efficiencies across sectors, this study applied the Logarithmic Mean Divisia Index (LMDI) decomposition method to analyze the drivers of carbon emissions in Gansu’s manufacturing sector, encompassing high, medium, and low-efficiency industries, and it identified vital factors affecting carbon emissions. A localized Long-range Energy Alternatives Planning System (LEAP) model for Gansu was also developed. This model includes six developmental scenarios to project future carbon emissions. The study results are as follows: (1) LMDI decomposition indicates that increased carbon emissions in the manufacturing industry primarily result from economic growth in less efficient sectors and the dominance of moderately efficient ones. (2) Under Optimization Scenario 6, a 50.82 × 10⁴ ton reduction in carbon emissions is projected for Gansu’s manufacturing sector by 2030 compared to 2020, marking the carbon peak. These outcomes provide valuable insights for policy reforms in Gansu’s manufacturing industry, aiming for carbon peaking by 2030.

1. Introduction

The manufacturing sector in China is a major energy consumer, accounting for 55.73% of the nation’s total energy usage [1] and contributing 56.93% of its carbon emissions [2,3], while also being a key driver of economic growth [4,5]. In response to the global climate crisis exacerbated by escalating carbon emissions [6,7], China has set the “30·60 target” to reach peak carbon emissions by 2030 and achieve carbon neutrality by 2060 [8,9]. Balancing economic growth with carbon reduction [10] is a formidable challenge for China’s development strategy. Therefore, understanding the factors driving carbon emissions in the manufacturing sector and designing peak carbon pathways under policy frameworks that reflect regional specifics are crucial for developing “dual carbon” strategies tailored to the manufacturing industry within these regions.

Exploring carbon emission reduction pathways is a viable strategy to meet set targets. The Logarithmic Mean Divisia Index (LMDI) method, a subset of Index Decomposition Analysis (IDA), is widely used to develop carbon-emission decomposition models [8]. These models are able to accurately identify and quantify various influential factors and their drivers [11,12,13]. Numerous studies have applied the LMDI method to assess factors impacting carbon emissions in the manufacturing sector [14,15,16]. These analyses typically categorize influencing factors into economic output, industrial and energy structures, and energy intensity [17,18]. Scholars often use a single metric, such as economic development level, average energy consumption, carbon emissions, or industry structure, to classify industries [19]. For instance, Jiang et al. applied the LMDI model to examine China’s carbon emissions from 2008 to 2019, assessing the contribution and impact of each factor. They discovered that increased regulatory intensity and industrial structure optimization help reduce emissions [20]. Similarly, Liu et al. used the LMDI model to identify critical factors affecting carbon emissions in Henan province, finding that expanding tertiary services and reducing the secondary industrial sector is crucial for emission control [21]. However, few studies consider different industries’ unique carbon emission characteristics and efficiencies.

Merely analyzing the drivers of carbon emissions in the manufacturing industry without projecting future emissions limits the depth of analysis regarding emission reduction potential and the generation of scientific and reasoned recommendations. The Stockholm Environment Institute and Boston University developed the Long-range Energy Alternatives Planning System (LEAP) model, a bottom-up energy-environment accounting tool for scenario analysis. This model, widely used in energy and carbon-emission studies [22,23,24], forecasts energy consumption based on various factors. Dong et al. used the LEAP model to predict carbon emissions for 12 major industrial sectors in Henan Province, identifying some sectors with potential for carbon neutrality [6]. Li et al. used the model to project energy demand and CO₂ emissions in the industrial sector of the Beijing-Tianjin-Hebei region, finding an increasing energy demand and a shift towards a cleaner energy structure [25]. Hernandez et al. applied the LEAP model to estimate CO₂ emissions in the industrial sector of the Bogotá region, highlighting the necessity of increasing electrical energy usage for emission reduction [26]. However, existing studies have primarily concentrated on overall industry, and research specifically addressing the manufacturing sector remains limited.

Gansu, a pivotal province in Western China for carbon-emission reduction, grapples with significant environmental challenges, fragile ecological systems, and a predominantly low-value-added industrial base [27]. Achieving a balance between economic growth and carbon emissions is crucial in Gansu. Addressing the gap in current research, this paper employs the LMDI-LEAP analytical framework to investigate carbon emission drivers in Gansu’s manufacturing sector and to model peak carbon pathways considering different carbon-emission-efficiency sectors and the latest state and provincial policy documents. We conducted a study on the manufacturing sector in Gansu Province to analyze the factors driving carbon emissions in high, medium, and low carbon-emission-efficiency industries. Through the decomposition analysis, we identified the key factors influencing carbon emissions, which were then integrated into the forecast to improve the accuracy of our results. Our study also suggests a scientific approach to reducing carbon emissions in the manufacturing industry in Gansu Province, which could be used as a reference by policymakers. This study focuses on the following crucial aspects: (1) Analyzing the decoupling status of manufacturing industries using the Tapio decoupling model. (2) Evaluating carbon-emission efficiency with the Super-SBM model. (3) Identifying factors influencing carbon emissions in Gansu’s manufacturing sector and outlining emission reduction strategies via the LMDI decomposition method. (4) Projecting the 2030 carbon emissions of Gansu’s manufacturing sector under various scenarios with the LEAP model, based on existing planning targets. In an optimal development scenario for 2030, the manufacturing sector in Gansu is projected to reach its carbon peak by that year. This research lays theoretical and empirical groundwork for the low-carbon development of Gansu’s manufacturing industry.

2. Materials and Methods

2.1. Tapio Decoupling

The OECD introduced the decoupling framework to characterize the relationship between energy consumption and economic growth. We employ the Tapio decoupling analysis method [28,29] to examine the link between carbon emissions and economic development:

$$e\left(C,G\right)=\frac{\frac{\Delta C}{C}}{\frac{\Delta G}{G}}$$

(1)

where $e\left(C,G\right)$ is the decoupling of carbon emissions and gross industrial product. $C$ and $G$ represent carbon emissions and total industrial output value at the beginning of the calculation period. $\Delta C$ and $\Delta G$ separately stand for the change in carbon emissions and the industry value added.

2.2. Super-SBM Model with Undesirable Outputs

We introduce a Super-SBM model [30,31,32] to account for undesirable outputs and assess carbon emission metrics across Gansu Province’s manufacturing industry and its 28 sub-industries.

Upon integrating unexpected outputs, the SBM model is represented by

$$\begin{array}{l}{\theta }^{*}=\underset{\lambda ,s^{-},s^{+}}{min}\frac{1+\frac{1}{m}{\sum }_{i=1}^m\frac{s_i^{-}}{x_{io}^t}}{1-\frac{1}{q+h}\left({\sum }_{r=1}^q\frac{s_r^{+}}{y_{ro}^t}+{\sum }_{k=1}^h\frac{s_k^{-}}{b_{ko}^t}\right)}\\ s.t.\left\{\begin{array}{l}{x}_{io}^t\ge {\displaystyle \sum _{t=1}^T}{\displaystyle \sum _{j=1,j\ne o}^n}{\lambda }_j^t{x}_{ij}^t-s_i^{-},i=1,2,\dots ,m;\\ y_{ro}^t\le {\displaystyle \sum _{t=1}^T}{\displaystyle \sum _{j=1,j\ne o}^n}{\lambda }_j^t{y}_{rj}^t+s_i^{+},r=1,2,\dots ,q;\\ b_{ko}^t\ge {\displaystyle \sum _{t=1}^T}{\displaystyle \sum _{j=1,j\ne o}^n}{\lambda }_j^t{b}_{kj}^t-s_k^{-},k=1,2,\dots ,h;\\ {\lambda }_j^t\ge 0\left(\forall j\right),s_i^{-}\ge 0\left(\forall i\right),s_r^{+}\ge 0\left(\forall r\right),s_k^{-}\ge 0\left(\forall k\right).\end{array}\right.\end{array}$$

(2)

where ${\theta }^{*}$ is the objective function, whose value can be more than 1; $x_{io}^t$, $y_{ro}^t$, and $b_{ko}^t$ represent the inputs, desirable outputs, and undesirable outputs; $s_i^{-}$ represents the shortage of good outputs; $s_r^{+}$ and $s_k^{-}$ correspond to excesses of inputs and undesirable outputs, respectively; and $\lambda$ is the non-negative intensity vector.

2.3. LMDI Decomposition

We utilize the extended Kaya identity with the LMDI decomposition method to dissect carbon-emission drivers and analyze their roles. We adopt the LMDI decomposition method’s additive approach [17,33], which breaks down carbon emissions according to

$$\Delta C=\Delta C_{eg}+\Delta C_{is}+\Delta C_{ei}+\Delta C_{es}+\Delta C_{ef}$$

(3)

where $\Delta C_{eg}$ represents the effect of economic growth, $\Delta C_{is}$ represents the industrial structure effects, $\Delta C_{ei}$ represents the energy intensity effect, $\Delta C_{es}$ represents energy structure effects, and $\Delta C_{ef}$ represents the effect of carbon-emission factors.

The five driving factors can also be expressed as

$$\Delta C_{eg}=\sum _{ij}C{L}_{ij}\cdot \mathit{ln}\left(\frac{e{g}^T}{e{g}^0}\right)$$

(4)

$$\Delta C_{is}=\sum _{ij}C{L}_{ij}\cdot \mathit{ln}\left(\frac{i{s}_i^T}{i{s}_i^0}\right)$$

(5)

$$\Delta C_{ei}=\sum _{ij}C{L}_{ij}\cdot \mathit{ln}\left(\frac{e{i}_i^T}{e{i}_i^0}\right)$$

(6)

$$\Delta C_{es}=\sum _{ij}C{L}_{ij}\cdot \mathit{ln}\left(\frac{e{s}_{ij}^T}{e{s}_{ij}^0}\right)$$

(7)

$$\Delta C_{ef}=\sum _{ij}C{L}_{ij}\cdot \mathit{ln}\left(\frac{e{f}_{ij}^T}{e{f}_{ij}^0}\right)$$

(8)

Here, $C{L}_{ij}=\frac{{C_{ij}}^T-{C_{ij}}^0}{{\mathit{ln}{C}_{ij}}^T-{\mathit{ln}{C}_{ij}}^0}$, ($C_{ij}^T\ne C_{ij}^T$).

2.4. LEAP Model

LEAP is a bottom-up energy-environment accounting tool for evaluating energy policies and addressing climate change. Scenario analysis, a crucial component of LEAP models, facilitates the projection of medium and long-term energy supply, energy transitions, final energy demand, and emissions of pollutant gases under various development scenarios. LEAP models employ predefined methods to simulate energy demand, supply transitions, and associated carbon emissions. The formula for computing energy demand in LEAP models is delineated as [22,23,26]

$$ED=\sum _i^3\sum _j^6{Al}_{ij}\times {Ei}_{ij}$$

(9)

where i represents three different sectors (i = 1, 2, 3), which refer to the high-efficiency industries, medium-efficiency industries, and low-efficiency industries, respectively; j denotes the type of final energy consumption, including coal, oil, natural gas, heat, electricity, and clean energy (j = 1, 2, 3, …, 6); Al denotes the activity level; and Ei represents the energy intensity.

2.5. Data Sources

Data on the gross industrial output for Gansu’s large-scale manufacturing from 2012 to 2021 is sourced from the Gansu Development Yearbook (2011–2020). The 2020 output value is derived by contrasting the 2020 growth rate with the 2019 data. For consistency, all years’ industrial output values are adjusted to the 2011 price level using Gansu’s factory price index.

Energy consumption data for Gansu’s manufacturing industry are drawn from the relevant section in China’s Provincial Energy Inventory 2011–2020, curated by CEADs (China Emission Accounts and Datasets, China Carbon Accounting Database).

Carbon emissions from the manufacturing industry are estimated following IPCC guidelines and parameters set by the Chinese government.

This study is based on the 31 manufacturing industries in the latest industrial classification for national economic activities (GB/T 4754-2017) for 2017 [34], combined with the changes and differences in the caliber of classification for the manufacturing industry in the data source during the study period, as well as merged similar industries. In this paper, we combined the textile wearing apparel and accessories industry with the leather, fur, feather, and related products and footwear industry; we also combined the rubber industry with the plastics industry. We merged the automobile industry with the railroad, ships, aerospace, and other transport equipment manufacturing industries, and we merged the comprehensive utilization of the resources waste industry with the metal products, machinery, and equipment repair industry. Therefore, this paper combined 31 manufacturing industries into 28 industries for the study. Table S1 lists the specific industry classifications.

3. Results

3.1. Status of Economy, Energy Consumption, and Carbon Emissions

3.1.1. Economic, Energy, and Carbon Emission Trends in Gansu Province’s Manufacturing Industry

Figure 1 presents the trends in economic performance, energy consumption, and carbon emissions in Gansu’s manufacturing industry from 2011 to 2020. Figure 1a illustrates an upward trend in manufacturing output from 2011 to 2014, followed by fluctuations from 2015 to 2020. The total industrial production of the manufacturing industry surged from CNY 475.038 billion in 2011 to CNY 714.967 billion in 2015, a 51% increase. This figure then declined to CNY 500.381 billion in 2019, a 30% decrease from 2015, and it rose to CNY 530.403 billion in 2020, a 12% increase from 2011.

The trends in energy consumption and carbon emissions largely mirror the pattern of industrial output in Figure 1b. During this period, Gansu’s manufacturing industry experienced an initial growth in energy consumption, followed by a decline and subsequent fluctuations. Specifically, in Figure 1c, Gansu’s manufacturing industry’s total energy consumption is 2737.69 × 10⁴ tce in 2011, rising to a peak of 3469.93 × 10⁴ tce in 2014, an increase of 27%. Subsequently, it decreased to 2866.95 × 10⁴ tce in 2016, a 17% reduction from 2014. From 2016 to 2020, energy consumption fluctuated, eventually increasing to 3091.47 × 10⁴ tce, up 13% from 2011. Analysis of energy consumption shows that coal and its derivatives comprised over 44% of Gansu’s energy use from 2011 to 2020, highlighting their dominance in the energy mix. The share of primary electricity and other energy sources increased from 7% in 2011 to 14% in 2020, reflecting the growth of new and renewable energy sources.

Figure 1d shows a 3% reduction in carbon emissions in Gansu’s manufacturing industry between 2011 and 2020. Secondary electricity and direct coal consumption comprised over 80% of these emissions. In 2011, Gansu Province’s manufacturing industry emitted 8393.45 × 10⁴ tons of carbon, peaking at 9708.44 × 10⁴ tons in 2014, a 16% increase. From 2014 to 2017, a downward trend was observed, with total carbon emissions dropping to 8124.80 × 10⁴ tons in 2017. Between 2017 and 2020, total carbon emissions fluctuated but remained relatively stable. In 2020, Gansu Province’s manufacturing industry’s total carbon emissions were 8129.92 × 10⁴ tons, approximately 3% lower than in 2011 and marginally better than a decade ago. Carbon emissions in Gansu’s manufacturing industry primarily stem from indirect emissions through secondary power use and direct emissions from coal and its derivatives.

Secondary electricity and direct coal consumption constitute over 80% of these emissions. In 2011, Gansu Province’s manufacturing industry’s total carbon emissions were 8393.45 × 10⁴ tons, rising to a peak of 9708.44 × 10⁴ tons in 2014, an increase of approximately 16%. From 2014 to 2017, a downward trend was observed, with total carbon emissions decreasing to 8124.80 × 10⁴ tons in 2017. Between 2017 and 2020, total carbon emissions fluctuated but remained relatively stable. In 2020, Gansu Province’s manufacturing industry emitted 8129.92 × 10⁴ tons of carbon, approximately 3% lower than in 2011 and a slight improvement from the level a decade ago. Most carbon emissions in Gansu’s manufacturing industry stem from indirect emissions through secondary electricity use and direct emissions from coal and its derivatives.

3.1.2. Economic, Energy, and Carbon Emission Trends in Gansu’s 28 Manufacturing Industries

Figure 2 shows the changes in energy consumption, carbon emissions, and economic efficiency for 28 manufacturing industries in Gansu Province from 2011 to 2020. The figures emphasize the structural changes within these industries. Significant disparities exist in energy consumption, carbon emissions, and economic efficiency among Gansu’s manufacturing industries. The Smelting and Pressing of Ferrous Metals (c18) industry, a specialist in ferrous metal processing, is the primary energy consumer and carbon-emission sector in Figure 2a,b, accounting for approximately 40% of the total energy consumption in the sector. Besides c18, the top energy-consuming industries are c19, c12, the Manufacture of Raw Chemical Materials and Chemical Products (c13), and the Manufacture of Non-Metallic Mineral Products (c17), with their 10-year energy consumption proportions being 20–30%, 16–18%, 7–11%, and 6–7%, respectively. These five industries are capital and resource-intensive. Additionally, the total energy consumption of other manufacturing industries accounts for only about 3% of the sector’s total energy consumption. Over the past decade, the industries of Smelting and Pressing of Non-Ferrous Metals (c19) and Processing of Oil, Coal, and Other Fuels (c12) have shown significant growth, collectively contributing to about 46% of the total industrial output in the manufacturing sector in Figure 2c.

3.2. Decoupling Analysis: Carbon Emissions and Total Industrial Output

Table 1. Decoupling efforts of economic growth and carbon emissions during 2011–2020.

Period	$\Delta \mathit{C}/\mathit{C}$	$\Delta \mathit{G}/\mathit{G}$	Decoupling Index	State
2011–2012	0.0836	0.1535	0.5445	Weak decoupling
2012–2013	0.0477	0.1895	0.2517	Weak decoupling
2013–2014	0.0189	0.0952	0.1981	Weak decoupling
2014–2015	−0.0573	0.0016	−35.7640	Strong decoupling
2015–2016	−0.0929	−0.0139	6.6841	Recessive decoupling
2016–2017	−0.0213	−0.1977	0.1077	Strong negative decoupling
2017–2018	0.0204	−0.0842	−0.2420	Strong negative decoupling
2018–2019	−0.0779	−0.0340	2.2884	Recessive decoupling
2019–2020	0.0635	0.0600	1.0579	Expansive coupling

outlines the decoupling states between overall carbon emissions and total industrial output in Gansu Province from 2011 to 2020. According to Table 1, the decoupling forms between carbon emissions and total industrial production in Gansu’s manufacturing industry evolved through phases of “weak decoupling, strong decoupling, recessive decoupling, strong negative decoupling, recessive decoupling, and expansive decoupling”. From 2011 to 2015, a weak decoupling state prevailed between the manufacturing industry’s carbon emissions and Gansu’s total industrial output, indicating relatively carbon-efficient growth. From 2016 to 2018, the relationship shifted between recessive and strong negative decoupling states, implying that carbon emissions remained high or increased despite a reduction in total industrial output. Between 2018 and 2020, the relationship reverted to recessive and expansive decoupling states, shifting from an imbalanced to a more harmonized trajectory in Gansu’s carbon emissions and total industrial output.

This study evaluates the decoupling relationships between carbon emissions and total industrial output in Gansu Province’s manufacturing industries, focusing on the periods of the 12th (2011–2015) and 13th (2016–2020) Five-Year Plans, as depicted in Figure 3. Following the 12th Five-Year Plan, industries c12, c18, c24 (Manufacture of Electrical Machinery and Equipment), and c28 (Comprehensive Utilization of Resource Waste and Metal Products, Machinery, and Equipment Repair) experienced poor negative decoupling between carbon emissions and industrial output, in contrast to other manufacturing industries, which exhibited either decoupling or linkage. The carbon emission-to-industrial output ratio in Gansu’s manufacturing industries during the 12th Five-Year Plan was favorable. Following the 13th Five-Year Plan, industries c4, c12, c15 (Manufacture of Chemical Fibers), c18, and c28 showed improved decoupling between carbon emissions and industrial output, while other manufacturing industries experienced negative decoupling or linkage. The carbon emission-to-industrial output ratio in Gansu’s manufacturing industries during the 13th Five-Year Plan period was less favorable than the 12th Five-Year Plan period.

Compared to the 12th Five-Year Plan period, the carbon emissions and total industrial output of the manufacturing industry in Gansu Province during the 13th Five-Year Plan period have shown a decline. This indicates that although the economy of most industries in Gansu Province’s manufacturing sector contracted during the 13th Five-Year Plan period, carbon emissions either decreased slowly or increased.

3.3. Decomposition Analysis

3.3.1. Carbon-Emission Efficiency Analysis

With the help of the Super SBM model, Figure 4 illustrates the calculated carbon-emission efficiency of Gansu Province’s manufacturing industry. Figure 4a shows that the carbon-emission efficiency of Gansu’s manufacturing industry initially increased, decreased, and then rose again. In 2020, the efficiency value climbed to 1.0258 from 0.7518 in 2011, marking a 36% increase. The average efficiency value over the decade stands at 0.8901, indicating an overall favorable trend. Gansu’s manufacturing industry’s carbon-emission efficiency exceeded 1 in 2013, 2014, 2016, and 2020. These years indicate a higher production level in Gansu’s manufacturing industry, with balanced capital, labor, and energy inputs. This condition contributed to sustainable economic growth, controlled carbon emissions, and yielded economic and environmental benefits. In 2018, the efficiency value dropped to an all-time low of 0.7345. A significant decline in manufacturing output within Gansu appeared in 2018, with higher energy consumption compared to the peak performance in 2016.

The carbon-emission efficiency data for the 28 industries show considerable variations within Gansu Province’s manufacturing sector. This study categorizes the carbon emission efficiencies of these 28 manufacturing industries to identify commonalities and differences. Detailed classifications are provided in Table S2. High-efficiency sectors primarily produce high-value-added products and form the backbone of the national economy. At the same time, low-investment technology-intensive and capital resource intensive traditional industries are distributed in medium and low-efficiency sectors, respectively. Figure 4b shows Gansu Province’s manufacturing industries’ average carbon-emission efficiency fluctuations from 2011 to 2020. Over the past decade, the average efficiency of high, medium, and low sectors and the overall industry exhibited a pattern of initial increase, followed by a decrease and a subsequent rise. Notably, the average efficiency of the high-efficiency sector exceeded that of the overall industry. Both medium and low-efficiency sectors demonstrated average efficiencies below the industry-wide average.

3.3.2. Decomposition of Carbon-Emission Factors

Employing the additive LMDI decomposition method, this study analyzed the effects on carbon emissions in Gansu Province’s manufacturing industry from 2011 to 2020, as shown in Figure 5. The examination of yearly factor effects and industry impacts revealed significant variations in the contributions of different factors and industries to carbon emission changes over time.

Figure 5a displays the cumulative impact of carbon emissions from 2011 to 2020, calculated from the summation of annual effects and their respective contribution rates during this period. Various factors exert distinct influences on carbon emissions. Figure 5a indicates that economic growth is the predominant factor driving the increase in carbon emissions in Gansu Province’s manufacturing industry. The economic growth effect contributed a cumulative total of 1306.72 × 10⁴ tons of carbon emissions from 2011 to 2020, accounting for approximately 37%. Economic growth positively affected high, medium, and low-efficiency industries; the primary cause of increased carbon emissions was in low-efficiency sectors, with a contribution rate of around 24%. The industry structure effect negatively influenced the overall carbon emissions in Gansu Province’s manufacturing industry, cumulatively contributing −542.12 × 10⁴ tons from 2011 to 2020, amounting to about −15%. In addition, the industry structure shows an adverse effect on high and low-efficiency industries. This is the most crucial reason for the decrease in the carbon emissions of the two categories of industries, with contribution rates of about −43% and −34%, respectively. In contrast, the industry structure positively affects medium-efficiency sectors. This is the most crucial reason for the increase in the carbon emissions of medium-efficiency industries, with a contribution rate of about 40%. The energy intensity effect positively impacted the overall carbon emissions in Gansu’s manufacturing sector, contributing 321.79 × 10⁴ tons from 2011 to 2020, accounting for about 9%. In addition, energy intensity positively affected high and low-efficiency industries. This is the cause of the increase in carbon emissions of the two categories of industries, with contribution rates of about 29% and 18%, respectively. In contrast, the energy intensity shows an adverse effect for medium-efficiency industries. This is the most crucial cause of the decrease in the carbon emissions of medium-efficiency industries, with contribution rates of about −22%. The energy structure effect was the most significant factor in reducing overall carbon emissions in Gansu’s manufacturing industry, cumulatively contributing −682.47 × 10⁴ tons from 2011 to 2020, with a contribution rate of approximately −19%. The energy structure shows an adverse effect for the three categories of high, medium, and low-efficiency industries, with a contribution rate of −13% for low-efficiency industries, −11% for medium-efficiency sectors, and an insignificant contribution rate of less than 1% for high-efficiency industries. The energy–carbon emission coefficient effect negatively influenced the overall carbon emission changes in Gansu’s manufacturing sector, cumulatively contributing −667.45 × 10⁴ tons from 2011 to 2020, with a contribution rate of about −19%. In addition, the energy structure exhibits an adverse effect for the three categories of high, medium, and low-efficiency industries, contributing −5%, −12%, and −11%, respectively.

Figure 5b presents the annual effects of each industrial category on carbon emissions in Gansu Province’s manufacturing sector from 2011 to 2020. The high-efficiency industry exhibited a negative impact from 2013 to 2019 and a positive effect in other periods, except for 2012–2013 and 2019–2020, where the results were the smallest and most stable among the three categories. The most significant contribution was in 2012–2013 (122.85 × 10⁴ tons), and the smallest was in 2016–2017 (−9.83 × 10⁴ tons), suggesting that high-efficiency industries are not the primary drivers of carbon-emission changes in the manufacturing industry of Gansu Province. The contribution of carbon-emission changes in 2016–2017 is the most minor (−9.83 × 10⁴ tons), indicating that high-efficiency industries are not the leading cause of carbon-emission changes in the manufacturing industry in Gansu Province. Medium-efficiency industries exhibited alternating positive and negative effects each year, with substantial variability in impact magnitude, showing no apparent pattern. In 2016–2017, the most significant change was a reduction of 1087.62 × 10⁴ tons; in 2019–2020, the most minor change was a decrease of 8.86×10⁴ tons. Low-efficiency industries showed positive effects from 2011 to 2014, with a mix of positive and negative impacts. The most significant changes occurred in 2014–2015 and 2017–2018. The most significant reduction in carbon emissions was 1689.56 × 10⁴ tons in 2015–2016, with the lowest being 966 × 10⁴ tons in 2017–2018. The smallest increase was 96.21 × 10⁴ tons, as seen in 2019–2020, suggesting that low-efficiency industries are a primary driver of carbon-emission changes in Gansu’s manufacturing industry.

3.4. Carbon Emissions Prediction

3.4.1. Key Factor Setting

The LMDI decomposition analysis reveals that the key factors affecting energy-related carbon emissions include the effects of economic development level, industry structure, and energy intensity. The impact of energy structure and carbon-emission factors are crucial in reducing carbon emissions.

As outlined in Gansu Province’s “14th Five-Year” Manufacturing Development Plan, aiming for an average annual growth of approximately 6.5% in additional manufacturing value, the target for the average yearly growth rate in total industrial output value from 2020 to 2025 is 6.5%. Based on the previous analysis of the economic growth and industry structure effects, and by prioritizing the development of high-efficiency industries while reducing medium and low-efficiency industries, the 2030 target for Gansu’s overall manufacturing industry structure is established. In line with the “Implementation Program of Carbon Peak in the Industrial Sector”, which sets a goal to decrease energy consumption per unit of added value by 13.5% by 2025 compared to 2020, the cumulative reduction rate of energy intensity for Gansu’s manufacturing industry is targeted at 13.5%. Following the energy development plan in Gansu Province’s “14th Five-Year Plan”, which proposes changes to the province’s energy consumption structure, the 2030 energy structure target for the manufacturing industry is established. Specific scenarios are detailed in

Table 2. Carbon emission scenario setting.

Scenario		2020	2030
			Scenario 1	Scenario 2	Scenario 3	Scenario 4	Scenario 5	Scenario 6
Average annual growth rate of total industrial output value (%)		-	6.50	6.50	6.50	6.50	6.50	6.50
Industry structure (%)	High-efficiency industries	32.59	32.59	36.00	32.59	32.59	36.00	40.00
	Medium-efficiency industries	42.58	42.58	41.00	42.58	42.58	41.00	40.00
	Low-efficiency industries	24.83	24.83	23.00	24.83	24.83	23.00	20.00
Cumulative decline rate of energy intensity (%)		-	0.00	0.00	13.50	0.00	13.50	25.00
Energy structure (%)	Coal and coal products	48.71	48.71	48.71	48.71	43.84	43.84	40.37
	Petroleum and petroleum products	9.00	9.00	9.00	9.00	8.55	8.55	8.55
	Natural Gas	3.59	3.59	3.59	3.59	5.74	5.74	7.15
	Heat	10.34	10.34	10.34	10.34	13.64	13.64	11.34
	Secondary electricity	14.35	14.35	14.35	14.35	13.64	13.64	11.34
	Primary electricity and other energy	14.01	14.01	14.01	14.01	17.89	17.89	21.22
Reduction rate of energy carbon emission coefficient (%)		-	0.00	0.00	0.00	0.00	0.00	13.00

.

3.4.2. Scenario Analysis

Figure 6 presents China’s projected primary energy demand under six distinct scenarios. The projections suggest a likely continuation of overall energy demand growth. This trend can be attributed to various factors such as economic growth, prioritization of high-efficiency sectors, reduced energy intensity, optimized energy mix, and decreasing energy–carbon emission factors.

By 2030, under Scenarios 1 through 6, the projected primary energy demand in China is expected to be 5803 × 10⁴ tons, 5616 × 10⁴ tons, 5157 × 10⁴ tons, 5803 × 10⁴ tons, 5009 × 10⁴ tons, and 4032 × 10⁴ tons, respectively. These amounts represent 1.88, 1.82, 1.67, 1.88, 1.62, and 1.30 times the level of 2020, respectively.

Table 3. Simulation results of cumulative effects of factors and carbon-emission changes in different scenarios (10⁴ t).

Scenario	ΔCeg	ΔCis	ΔCei	ΔCes	ΔCef	ΔC
Scenario 1	7130.96	0.00	0.00	0.00	0.00	7130.96
Scenario 2	7002.32	−357.51	0.00	0.00	0.00	6644.81
Scenario 3	6689.45	−9.7 × 10−13	−1540.53	0.00	0.00	5148.93
Scenario 4	6931.39	−1.05 × 10−12	0.00	−547.64	0.00	6383.75
Scenario 5	6405.58	−300.90	−1475.15	−492.22	0.00	4137.32
Scenario 6	5088.26	−642.40	−2324.42	−1047.04	−1125.21	−50.82

presents the projected carbon emissions under six different scenarios. Figure 7 shows the contribution of the cumulative effect of future carbon emissions from the manufacturing industry in Gansu Province. In Scenario 1, considering only the economic growth factor, the carbon emissions of Gansu Province’s manufacturing industry are projected to increase by 7130.96 × 10⁴ tons by 2030. Based on the economic growth rate established by current planning targets, carbon emissions are expected to rise by approximately 88% by 2030, compared to 8129.92 × 10⁴ tons in 2020. Scenario 2 incorporates adjustments to the structure of high, medium, and low-efficiency industries and economic growth. In the future, the manufacturing industry in Gansu Province will be prioritized according to the development of high-efficiency sectors, i.e., the share of high-efficiency industries will grow by 3%. Under this scenario, carbon emissions from Gansu’s manufacturing industry are projected to increase by 6644.81 × 10⁴ tons by 2030, a reduction of 486.15 × 10⁴ tons compared to Scenario 1. Adjusting the industrial structure has a specific effect on controlling carbon emissions. Scenario 3, while considering economic growth, also includes a reduction in energy intensity. The energy intensity of Gansu’s manufacturing industry is projected to decrease by 14% by 2030 compared to 2020. Under this scenario, the carbon emission of the Gansu Province manufacturing industry will increase by 5148.93 × 10⁴ tons by 2030. There is a significant effect on reducing carbon emissions by reducing energy intensity. Scenario 4 focuses on optimizing the energy structure and economic growth. Carbon emissions from Gansu Province’s manufacturing industry will increase by 6383.75 × 10⁴ tons by 2030. Optimizing the energy structure has a specific effect on suppressing the growth of carbon emissions. Scenario 5 involves adjustments to industry structure, energy intensity, and energy structure alongside economic growth. Under this scenario, the carbon emissions from the manufacturing industry in Gansu Province will only increase by 4137.32 × 10⁴ tons by 2030, which is a decrease of 2993.64 × 10⁴ tons, or about 75%, compared with the carbon emissions from Scenario 1. Scenario 6 implements comprehensive measures on top of economic growth, including adjustments to industry structure, energy intensity, energy structure, and the energy–carbon emission coefficient. In this scenario, the carbon emissions of the manufacturing industry in Gansu Province will be reduced by 50.82 × 10⁴ tons in 2030 compared with the carbon emissions in 2020, which will achieve the carbon peak target. Comprehensive measures have the most significant effect on carbon reduction.

4. Discussion

4.1. Status of Carbon Emissions and Total Industrial Output

Gansu Province’s manufacturing industry experienced notable growth throughout the study period, driven by the “strong industry province” strategy. Influenced by policy directives from the 19th National Congress of the Communist Party of China, Gansu shifted its focus towards balancing ecological conservation with economic growth. Constraints in manufacturing and ongoing structural adjustments led to a curtailment in industrial growth, resulting in decreased total output. However, development levels rebounded in 2020. From 2011 to 2020, capital and resource-intensive industries in Gansu demonstrated strong growth. Notably, these industries are characterized by high energy consumption and carbon emissions. Research [35] suggests that Gansu’s manufacturing sector is predominantly industrialized but maintains a low structural level.

In 2014, carbon emissions in Gansu reached a peak of 9708.44 × 10⁴ tons, marking a 16% surge, mainly due to the expansion of the manufacturing industry [36]. During the “12th” and “13th Five-Year Plans”, Gansu enhanced its manufacturing sector while strategically eliminating obsolete production capacities. As a result, the manufacturing industry experienced a green transformation, significantly reducing energy consumption. This transformation likely contributed to the industry’s slowdown or decrease in energy consumption after 2014. Figure 1 and Figure 2 emphasize coal and its derivatives as the primary energy sources. The heavy reliance on these sources may hinder carbon-reduction efforts [37]. Notably, there was a general decline in coal usage during this period, interspersed with minor increases. Between 2011 and 2020, the proportion of primary electricity and other energy types in Gansu’s manufacturing industry increased significantly from 7% to 14%. This shift indicates a move towards cleaner energy in Gansu’s manufacturing sector, promoting a low-carbon and more sustainable energy infrastructure. Such structural improvements can reduce reliance on fossil fuels and mitigate carbon emissions [38].

The “12th Five-Year Plan” emphasizes the “transfer and upgrade of traditional manufacturing” as a critical strategy, aiming at the transformation and green evolution of the manufacturing industry. Figure 3 illustrates the decoupling trend between carbon emissions and energy consumption across the 12th and 13th “Five-Year Plan” periods [39]. In the past decade, Gansu’s manufacturing industry has increasingly utilized primary power and alternative clean energy sources, effectively reducing carbon emissions per unit of energy. Maintaining this downward trend will enhance efforts to curtail carbon emissions and foster a low-carbon economic pathway.

4.2. Carbon-Emission Efficiency

In 2013, 2014, 2016, and 2020, Gansu Province’s manufacturing industry surpassed the production frontier, achieving carbon-emission efficiency values exceeding 1. Throughout these years, Gansu’s manufacturing industry demonstrated balanced capital, labor, and energy inputs. This sector achieved economic growth with effective control over carbon emissions, leading to an efficient and low-carbon economy. In 2016, the carbon emission efficiency reached 1.0463. In 2016, Gansu’s manufacturing industry achieved substantial industrial output with minimal energy consumption and carbon emissions. The combination of high economic efficiency, lower energy consumption, and reduced emissions significantly contributed to the peak efficiency value in 2016. In other years, Gansu’s manufacturing industry consistently reported carbon-emission efficiencies below 1, falling behind the production frontier. Specifically, efficiency dropped to a historical low of 0.7345 in 2018. In contrast to 2016, 2018 experienced a significant decrease in output and increased energy consumption in Gansu’s manufacturing sector, indicating a less optimized development approach. After 2018, efficiency rebounded, exceeding 1 by 2020, signaling an effective recovery. For the future growth of Gansu’s manufacturing industry, prioritizing carbon-emission efficiency could yield better outcomes than solely focusing on reducing fixed investments and workforce.

From 2011 to 2020, the average carbon-emission efficiency of Gansu Province’s manufacturing industry was 0.1804. There is a pronounced disparity in carbon-emission efficiency among Gansu Province’s manufacturing industries. Among the 28 manufacturing industries, industry c4 stands out with a commendable efficiency value of 0.8470, significantly surpassing its counterparts. This indicates consistent top-tier carbon-emission efficiency in industry c4. Sectors c8, c10, c9, c17, c13, and c27 show below-par carbon-emission efficiencies, averaging below 0.1, indicating significant potential for improvement. This highlights the latent potential to enhance efficiency across Gansu’s manufacturing industry. Therefore, improving carbon emission efficiency, especially in underperforming sectors, is crucial for sustainable energy conservation and promoting a greener trajectory in Gansu’s manufacturing sector.

4.3. Carbon-Emission Factors

Different factors have influenced changes in carbon emissions during various phases. At the start of the “12th Five-Year Plan”, economic growth had a significant positive impact, in contrast to the considerable adverse effect of energy intensity. In the concluding phase of the “12th Five-Year Plan” and the beginning of the “13th Five-Year Plan”, the industry’s structure played a crucial role in reducing carbon emissions. Throughout the “13th Five-Year Plan”, no single factor predominantly influenced emission variations. Economic growth and energy intensity also play pivotal roles in emission fluctuations. Future manufacturing strategies should prioritize these aspects to ensure economic growth and energy intensity do not unduly escalate carbon emissions.

From an industry perspective, various sectors demonstrate different contribution levels to carbon emissions across distinct timeframes. The high-efficiency sector predominantly shows negative impacts with minimal assistance to carbon emissions. The medium-efficiency sector fluctuates between positive and negative effects, experiencing significant shifts in contributions without a discernible pattern. The low-efficiency sector alternates between positive and negative influences, showing substantial contribution variations. This sector remains the primary contributor to shifts in carbon emissions within Gansu’s manufacturing industry.

In summary, various factors and industry classifications impacted carbon-emission changes in Gansu’s manufacturing sector from 2011 to 2020. In terms of factors, both economic growth and energy intensity have contributed to increases in carbon emissions in the manufacturing sector. Regarding industry classifications, high and medium-efficiency sectors have exacerbated carbon emissions, while low-efficiency categories have mitigated this increase. A detailed analysis of influencing factors and industry contributions offers valuable insights into strategies for reducing carbon emissions in Gansu’s manufacturing industry. These strategies include structured development of the low-efficiency sector, avoiding unnecessary expansion. The objectives encompass optimizing energy usage, reducing the scale of the medium-efficiency industry, and lowering energy intensity in both high and low-efficiency sectors. These goals can be achieved by increasing low-carbon energy use and adopting technological innovations that reduce carbon emissions.

It is found that the effect of the same factor on the carbon emissions of industry and the manufacturing industry varies in different study areas when comparing against the results for other regions. The industry structure shows a negative effect for Gansu Province and a positive effect for China [40], indicating that the adjustment and optimization of the industry structure in Gansu Province is better and inhibits the increase of carbon emissions; energy intensity shows a negative effect for Jiangsu Province [41] and Henan Province [6] and a positive effect for Gansu Province, indicating that the energy intensity of the manufacturing industry in Gansu Province is still relatively high and has not yet reached the level of other regions; therefore, lowering the energy intensity is the focus of carbon-emission reduction for the manufacturing industry in Gansu Province in the future. Therefore, reducing energy intensity is the focus of carbon-emission reductions in the manufacturing industry in Gansu Province in the future; energy structure has a negative effect on Gansu Province and Henan Province, indicating that the energy structure of the manufacturing industry in Gansu Province and Henan Province is better and reduces carbon emissions; the source carbon-emission coefficient has a negative effect for Gansu Province and China [42] and inhibits the increase of carbon emissions.

4.4. Future Carbon Emissions and Policy Suggestions

Against the backdrop of ongoing economic growth, a comparison of the simulated carbon emission from Gansu Province’s manufacturing industry under six scenarios reveals that, by 2030, carbon emissions will increase under all scenarios, albeit to varying degrees. The effectiveness of different measures in reducing carbon emissions varies, with the most to least effective being Scenario 6, Scenario 5, Scenario 3, Scenario 4, Scenario 2, and Scenario 1. Scenario 6 applies a combination of lowering energy–carbon emission coefficients, reducing energy intensity, optimizing energy structure, and prioritizing the development of high-efficiency industries. This comprehensive approach constitutes the optimal scenario, potentially enabling carbon emissions to peak in 2030.

Based on our primary findings, we suggest the subsequent policy directives to further the agenda of carbon-emission reduction.

(1)

There is an imperative to uphold delineated policies, emphasizing those championing low-carbon development. Such approaches require meticulous formulation. The strategies for enhancing carbon efficiency and curtailing emissions vary across high, medium, and low-efficiency manufacturing sectors. Consequently, it is incumbent upon policymakers to devise tailored measures that augment energy efficiency and curtail emissions across diverse manufacturing domains. Such interventions should account for each industrial segment’s idiosyncratic developmental traits and underlying principles.

(2)

The interplay between economic growth, energy savings, and carbon mitigation should be balanced, prioritizing enhancing the carbon-emission efficiency. A highly efficient sector which significantly bolsters the national economy should be supported in its production prowess and product valuation, whilst maintaining energy conservation and minimizing its carbon footprint. This approach will amplify the economic yields. For the technology-driven medium-efficiency sector, ramping up investments in R&D, energy conservation, and environmental stewardship is pivotal for growth augmentation. Such an approach can cultivate a robust industrial ecosystem, translating assets into tangible economic advantages. There is a pressing need to swiftly phase out obsolete production capacities, enabling a shift in the paradigm for capital and resource-heavy industries that currently demonstrate low efficiency. Technological advancements can help us transition to a more eco-friendly and low-carbon way of life. The goal is to usher in a developmental paradigm marked by minimal energy usage, reduced pollution, diminished emissions, and optimized efficiency.

(3)

The adoption of clean energy should be advocated while refining the energy consumption matrix. In Gansu Province, the manufacturing industry primarily relies on carbon-intensive fossil fuels, while clean energy sources, like wind and photovoltaic, remain limited. Executing the “14th Five-Year Plan for Energy Development in Gansu Province” is crucial. This plan advocates for the broadened development and adoption of renewable energies, reducing costs for technologies such as wind and photovoltaics and bolstering their prevalence. Additionally, relevant authorities must craft policies emphasizing clean energy use. These policies should encourage manufacturers to shift away from fossil fuels, increase the contribution of renewables to the overall energy mix, and enhance the prevailing energy consumption model.

(4)

High-quality development should be pursued to ensure steady economic progression and refine the industrial framework. Firstly, it is vital to establish pragmatic development targets for Gansu’s manufacturing industry and shun thoughtless scaling. Emphasis should be on curbing or phasing out industries characterized by high energy consumption, substantial emissions, and significant pollution. Secondly, it is essential to fine-tune the structure of the manufacturing industry. Prioritizing initiatives with high efficiency and significant added value is crucial. By adhering to the “14th Five-Year Plan” for Gansu’s manufacturing industry, we should foster the growth of energy-efficient, low-emission, technologically advanced, and eco-friendly manufacturing. This condition also necessitates a balanced approach between traditional core industries and new emerging ones.

4.5. Limitations and Further Perspectives

Several important considerations should be noted. Firstly, due to inherent delays in releasing official data, this study was limited to data available up to 2020. Additionally, to avoid issues from extended data gaps and changes in statistical methods, the research timeframe was confined to 2011–2020. Considering the use of data after lifting COVID restrictions, future studies could incorporate data after industrial production had returned to normal, and in-depth long-term time-series analyses could be carried out by effectively merging and standardizing earlier datasets. Moreover, although the current development objectives extend only to 2025, the limited constraints restrict comprehensive scenario modeling. Future research could utilize the development experience of similar areas (domestically and internationally) in Gansu Province. This could broaden the scope of scenario settings, leading to a comprehensive and systematic analysis of the factors that affect the future development of carbon emissions.

5. Conclusions

This study examines the entire manufacturing industry of Gansu Province, covering 28 specific sectors, such as non-ferrous and ferrous metal smelting and rolling processing. Initially, the study explores decoupling dynamics in Gansu by assessing the relationship between carbon emissions and industrial output. Subsequently, a Super-SBM model, including undesirable outcomes, is used to measure the carbon-emission efficiency of Gansu’s manufacturing industry and its 28 constituent sectors. The LMDI model is applied to analyze the impact of various factors on emissions, including economic growth, industrial structure, energy intensity, energy composition, and carbon-emission metrics. Finally, multiple scenarios were designed using the LEAP model to predict carbon emissions in Gansu Province. The main conclusions of this study are as follows:

(1)

From 2011 to 2020, Gansu’s manufacturing industry experienced trends that mirrored each other in terms of overall economy, energy consumption, and carbon emissions: an initial rise followed by a decline and rebound.

(2)

Gansu Province’s manufacturing sector generally demonstrates commendable carbon-emission efficiency. Between 2011 and 2020, the average carbon-emission efficiency of the manufacturing industry was 0.8901. However, significant disparities exist in carbon-emission efficiency among the 28 manufacturing industries, with an average efficiency of 0.1804 across these sectors. Industry-specific characteristics primarily influence the carbon-emission efficiency in these industries.

(3)

The increase in carbon emissions in Gansu’s manufacturing sector can be primarily attributed to the economic growth of low-efficiency industries and the significant presence of medium-efficiency industries.

(4)

The simulations revealed varying degrees of increase in carbon emissions across these scenarios. Notably, under Scenario 6, carbon emissions are projected to decrease by 50.82 × 10⁴ tons compared to 2020, potentially achieving the carbon peak target. These results underscore those strategies, such as restructuring the industrial framework, reducing energy intensity, optimizing the energy matrix, and lowering the energy–carbon emission coefficient, effectively reducing carbon emissions.