Introduction

Being one of energy-intensive industries in China, the smelting and pressing of nonferrous metals industry (SPNFMI) consumed 25.56 million tons of standard coal equivalent in 1992 to 207.07 million tons in 2015 and produced nonferrous metals from 3.43 to 51.56 million tons1, releasing greenhouse gases (GHGs) and “three wastes” (waste gas, wastewater and waste solids). The rapid growth of the industry poses a threat to its sustainable development. It is necessary to evaluate resource consumption and ecological impacts for ensuring sustainable development of this expanding industry. Studies focused on the industry aspects: such as energy consumption2 and energy savings3,4; emissions5,6,7,8,9; currency10,11,12,13; and technological progress14,15. However, these researches did not reflect the resource degradation from a sustainable development perspective.

To establish an integrated measurement method from a sustainability perspective, a holistic view of processes should be adopted16. Being an alternative to other analysis approaches, exergy analysis could reflect the resource utilization situation and environmental emissions with a solid foundation, and locate the potential to improve the sustainability of the system. A higher exergetic value represents a greater ability of resources to influence the environment17,18. Exergy is a meaningful instrument to appraise the usefulness and degradation of various resources with scientific basis19,20 and emissions affect the environment21. Chen integrated exergy emissions impact for the remediation of “three wastes”16. Subsequently, more environmental emissions comprising greenhouse gases (GHGs) and “three wastes” in China were addressed22,23,24.

Scholars have discussed the exergy flux in industries and nations to enhance resource transformation efficiency. Industries have been considered: transportation sector in China25,26,27,28, Saudi Arabia29, Jordan30, Turkey31,32,33, Greece34; agriculture in China35,36,37, Malaysia38, Saudi Arabia39, OECD40; damaskrose production41; the public and private sector in Saudi Arabia42; commerce in Malaysia43; energy sector in Mexico44 and energy-intensive industries in Denmark45. Countries or regions, such as: Europe (Spain46, the UK47, regions of Denmark48, Greece49 and all of Europe50, America (the United States51,52, regions of Canada53, Latin America54, Colombia55 and São Paulo State56, and Asia (India57, Japan58, Singapore59 and China16,60,61,62,63,64 have been included in this process. Then researches have been broadened to contrast exergy efficiency in global level65,66,67,68,69. These studies are classified into three categories: concentrating on energy utilization Saudi Arabia39; material and energy in exergy accounting for in Norway70 and China62; labour and currency being included into the exergy accounting for Norway21, Italy71,72, Turkey73 and China16,74,75 .

These studies show that sustainable development from the exergy perspective displays the potential to raise the exergy conversion efficiency76. Exergy analyses achieve sustainable development by estimating the types of resources and by improving the transformation efficiency in a thermodynamic view. Extended exergy accounting (EEA) is a broadening exergy analysis, containing energy, materials, labour, capital, and environmental remediation measured in joules. It is valid to compare different inputs and outputs by EEA based on thermodynamics law77. Furthermore, the evaluation approach of natural, social and environmental effects could be regarded as the ‘‘ecological cost’’ of resources to system. EEA is the instrument to connect the thermodynamic, ecological costs and resource depletion quantified by exergy74,78.

With huge population, upgrading life level and urbanization, large quantities of nonferrous metals consumption is the solid fact that China needs to face. Researchers performed on exergy calculation in China containing the SPNFMI in single year17,60,74, not concentrating on the evolutionary characteristic of this energy-intensive industry. Exergy research focused on the metal industry covering both ferrous and nonferrous metals79 and exergy of emissions accounting based on the expenditure for environmental remediation74, not the exergy of different emissions and excluding the GHGs. Previous studies did not reflect the characteristics, the evolution of the structure and sustainable development indicators of the SPNFMI. The physical sustainability of the SPNFMI remains to be revealed systematically.

To improve the resource utilization level of the SPNFMI and realize its sustainable development, EEA was employed to analyse the energy-intensive industry during 1992–2015. The inputs and outputs of the SPNFMI were estimated based on the EEA. Indicators representing the sustainability of the industry were calculated, then suggestions were provided to promote the sustainable development level of the SPNFMI. The results offer reliable information to optimize resource use80, with essential implications for sustainable development.

Methodology and data

With reference to previous research17,60,81, metal production of aluminium, lead, copper, zinc, tin, and nickel were selected to represent the SPNFMI, although several metals are included. In EEA model, materials, energy carriers, capital, labour and emissions are transformed into joules26,82. Based on previous research74,78,the EEA could be shown as:

$$EE\,=\,{E_E}+{E_M}+{E_K}+{E_L}+{E_R}$$
(1)

EM and EE are the ores of nonferrous metals, scrap nonferrous metals and energy carriers (coal, electricity, coke, oils and natural gas) input in the industry. EL, EK, ER and Yields are the exergy equivalents of human labour, capital, environmental remediation costs and metal production. To account for the exergy of resources, the Earth’s ecosystem was selected as a standard environment22. Compared with those of imported ores, the exergies of lower-grade copper, aluminium, lead, tin, zinc and nickel extracted in China are estimated as 0.026, 0.3, 0.021, 0.02, 0.046 and 0.04 PJ/Mt (1015J/million tons), perspectively17,60. With regards to the high grade imported ores, the exergies of copper ore and bauxite were estimated as 1.1 and 2.0 PJ/Mt (1015J/million tons), respectively81. The thermophysical exergy of ores, energy and metals is ignored22,60. Different materials and energy were transformed into joules by their transformation factors provided by previous studies78,83. Detailed information is listed in the supplementary Table 1. According to former studies74,78, EL and EK are quantified as follows:

$${E_L}={\text{ }}f \times {e_{surv}} \times {N_h}$$
(2)
$$\:E_{K}=\frac{M2}{s{\times\:N}_{w}\times\:W}\times\:E_{L}$$
(3)

where f indicators a correction factor related to the standard of life in countries or areas (f = HDI/HDI0, HDI is the human development index displayed from the United Nations); esurv is the 107 J/(person×day) implying the essential exergy consumption; Nh represents the population in China; M2 is the money stock (unlike the Western banking systems, large proportion of M2 is deposits in China. Thus, we accept the GDP as the indicator74,78; Nw and s are the number of workers and the average wage, individually; W symbols the average workload 2000 h16,78. The boundaries of capital and labour were limited in the SPNFMI of China. Based on previous research26,74,78, exergy equivalents of human labour and exergy equivalents of capital in China were calculated. As part of the total exergy equivalents of capital in China, EK of manufacturing industry was obtained based on its Nw and s. Then EL of manufacturing industry was calculated by transposing the Eq. (3). In the same way, EK and EL of the SPNFMI were evaluated.

Emissions comprise two parts: “three wastes” (waste gas, solid waste and wastewater) and GHG emissions. The exergetic values of the main GHGs (CO₂ and N₂O) are 0.45 and 2.4 PJ/Mt (1015J/million tons) in SPNFMI, respectively22. Considering previous research, wastewater includes the Chemical Oxygen Demand (COD) and ammonia; Waste gas comprises SO2, NOx, soot and dust.

It is necessary to assess the sustainable development of society through measurement indicators84. Considering the former research76, indices are applied to reflect the evolution of the industry in exergy perspective.

$$RECC\,=\,Yields/\left( {{E_E}+{E_M}} \right)$$
(4)
$$RLCR=\left( {{E_L}+{E_K}} \right)/\left( {{E_E}+{E_M}} \right)$$
(5)
$$EECC\,=\,Yields/\left( {{E_E}+{E_M}+{E_L}+{E_K}+{E_R}} \right)$$
(6)
$$ERR=\left( {{E_L}+{E_K}} \right)/\left( {{E_E}+{E_M}+{E_R}} \right)$$
(7)
$$EEER={\text{ }}\left( {{E_E}+{E_M}} \right)/\left( {{E_E}+{E_M}+{E_L}+{E_K}+{E_R}} \right)$$
(8)

RECC is the ratio of Yields to the natural resource consumption meaning the resource exergy conversion coefficient without the externalities. A higher RECC value shows more sustainable utilization of the industry. RLCR is the capacity of resource depletion to generate economic benefit, and is defined as the proportion of EL and EK to EE and EM. The decline of RLCR value means less labour and capital corresponding to one unit natural resource consumption. The EECC means the extended exergy conversion coefficient, implying the ability of extended exergy utilized for metals production. With higher EECC value the system has lower exergy loss ratio, creating more outputs with less resource inputs. The ERR means the intensity of economic action in the industry. A higher ERR value shows more economic action corresponding to one unit natural resource consumption and emission. EEER is the ratio of EE and EM to the extended exergy. Bigger EEER value reflects lower socio-economic-environment influences in the resource transformation process. The workflow of this study is detailed in Fig. 1.

Since data were unavailable, we chose the survey period from 1992 to 2015. The data originated from The Yearbook of Nonferrous Metals Industry of China1, China Statistical Yearbook85, China Environment Yearbook86 and China Labor Statistical Yearbook87.

Fig. 1
figure 1

Methodological flow chart.

Full size image

Results

Extended exergy variations

Figure 2a describes the variations of the extended exergy constituents in the SPNFMI from 1992 to 2015. The total extended exergy increased from 0.77 to 6.36 EJ an increase of 7.3 times. The only negative growth occurred in 1998, at 0.39% and the fastest growth reached 32.24% in 2010. The exergetic values of energy, material, labour, capital and environmental remediation increased 11.71, 37.50, 0.29, 0.17 and 9.62 times, respectively. The exergy of labour and capital notably lagged behind the rate of the total EEA value. The exergy of energy and material accounted for the highest proportion of the EEA from 58.20% to 91.60% with changes.

In the exergy of labour and capital accounting, the values in the manufacturing sector in China increased 0.12 and 3.89 times, respectively, from 2000 to 2007 78, with the same growth trend as that of the SPNFMI. The nonferrous metals yields made progress and labour did not increase proportionally, showing the productivity of the SPNFMI increased notably. With the urbanization, infrastructure construction and the upgrading of the consumption level in China, the requirement for nonferrous metals does exist. Given China’s large population and relative resource shortage, improving resource utilization efficiency represents a reasonable way to achieve sustainable development of SPNFMI. Therefore, the key factor is not labour and capital, but rather technical progress. Yield per unit of labour increased steadily from 0.15 to 2.90 in the period. The ratio of yield and emissions increased from 3.56 to 8.29, it grew in the initial years and declined in the last few years. They showed that technical progress was of great significance to the SPNFMI production and similar conclusions were concluded by other studies14,15,88.

The proportion of EE in EEA nearly consistently increased from 57.39% to 87.84% from 1992 to 2015, except in 1995, 1999 and 2000 as shown in Fig. 2b. The percentage of EM ranged from of 0.81% to 3.76% over the period with fluctuations, reaching the highest value of 6.67% in 2007. The proportions of EL and EK clearly decreased from 40.07% to 0.03%, respectively, to 6.22% and 0.004%, with slight fluctuations. The proportion of ER remained stable relatively, ranging from 1% to 2% from 1992 to 2014, and reaching 2.17% in 2015.

Fig. 2
figure 2

Extended exergy inputs. (a) Amount of extended exergy inputs. (b) Proportion of extended exergy inputs.

Full size image

Exergy equivalent of energy and material

With its leading position, EE increased from 0.44 to 5.58 EJ. Figure 3 displays that the growth rate of EE in the SPNFMI increased much faster than that of the Chinese industry after 2001, especially in 2010. It fluctuated with the lowest value of -2.29% in 2000, and reaching values of 40.50% and 23.00% in 2010 and 2015, respectively, compared to the Chinese industry’s figures of 8.24% and − 1.78% for the same years. The proportion of EE input in the SPNFMI accounting for the total Chinese industry increased from 1.32% to 3.91%. Energy consumption of the SPNFMI increased significantly from the exergy perspective, which differs from previous finding89 in which the energy consumption rate showed a declining trend. Coal and electricity increased from 0.24 to 0.11 EJ (54.30% and 26.12% of the energy input) to 3.21 and 1.98 EJ (57.56% and 35.50%). The ratio of coal decreased during the initial years then increased. Electricity generation relies primarily on thermal power in China. Therefore, the main energy carrier is coal.

Fig. 3
figure 3

Energy exergy inputs.

Full size image

EM involving ores and scrap metals, expanded from 6.21 PJ to 238.98 PJ with drastic changes and its growth rate declined significantly in 1996, 1998, 2001, and 2015 shown in Fig. 4. The growth rate of EM reached 83.33% in 1995 which was due mainly to net imported scrap aluminium (11.17 PJ) contrasting EM value was 10.50 PJ in 1994. The lowest value was − 22.64% in 2001 since net imported scrap aluminium reduced from 26.22 PJ in 2000 to 11.84 PJ in 2001. The dominant part was always aluminium, accounting for 52.12% in 1992 and 86.94% in 2015. The proportion of copper declined with changes from 37.65% to 11.34%. Zinc, lead, tin and nickel took the third, fourth, fifth and sixth position accounting for 10.23% of EM in 1992 and 1.72% in 2015, and nickel took the place of the zinc from 2010.

As shown in Fig. 4, the material exergy inputs contained three parts: domestic, imported and scrap metals enhancing from 2.48, 1.68 and 2.05 PJ to 25.50, 137.15 and 76.33 PJ, increasing 9.30, 80.87 and 37.11 times, respectively. The exergy of exported ores and scrap metals was much lower than that of imported materials. The export ores were stable from 0.30 to 0.61 PJ. The exergy value of exported scrap metals was much lower than that of imported scrap metals. The exergy value of domestic ores was lower than that of imported ores. Imported ores and the scrap metals were both much more than that of domestic ores. Since 2007, the imported ores increased significantly and took more proportion of the materials. Though domestic ores increased its proportion became small compared to the imported ores and scrap metals. The ratio of energy to materials significantly declined from 70.7 in 1992 to 23.4 in 2015 with fluctuations. The highest value was 82.1 in 1993 and the lowest value was 11.1 in 2007. The amount of imported material climbed sharply and scrap might be a better choice for the SPNFMI in the future.

Fig. 4
figure 4

Material exergy inputs.

Full size image

Exergy equivalent of externalities

The values of EL and EK in the SPNFMI increased by 128.90% and 117.40% during the study period. The EL and EK of the total Chinese manufacturing industry grew 159.62% and 145.38% in the same period. The conclusion has the same tendency in previous research from 2000 to 2007 78 and differ from those observed from 2012 to 2020 76. The highest growth rates of EL and EK were 10.51% in 1994 and 11.00% in 2007, respectively. Prominent negative growth rates of EL and EK was 9.34% and 13.96% both in 2002. The sum of EL and EK grew in the accounting period, however, considering the increase in the total EEA, the proportion of the EL and EK in EEA declined significantly, meaning that the labour and capital were not as critical as before.

In the SPNFMI, the GHG emissions included mainly carbon dioxide and nitrous oxide from energy consumption. The emission factors of carbon dioxide from coal, oil and natural gas are 1.7 kg/kg, 3.0 kg/kg and 2.16 kg/m3, whereas nitrous oxide from those energy carries are 1.4, 0.6 and 0.1 kg/TJ, respectively22. CO2 emissions expanded from 9.68 PJ in 1992 to 116.36 PJ in 2015, increasing 11.02 times, coal being the largest source (85.06% to 95.32% of the CO2 emissions), the proportion of oil declining (from 14.94% to 1.13%), and the proportion of gas increased (from 0% to 3.55%). N2O emissions were 0.82 TJ in 1992, with 92.10% from coal and 7.90% from oil, and it expanded to 10.60 TJ in 2015 increasing 11.98 times, with 95.6% from coal, 0.50% from oil and 3.80% from natural gas. Coal was also the largest contributor to N2O emissions. The fraction of natural gas increased and the oil contribution decreased. Exergy of N2O was less than four magnitudes than that of CO2, therefore, the total greenhouse gas emissions were more in line with the emissions of CO2.

The annual emissions of the “three wastes” in the SPNFMI increased from 3.33 PJ in 1992–8.83 PJ in 2014 with fluctuations and swelled to 21.83 PJ in 2015. The growth rate of the “three wastes” decreased in 1998, 1999, 2002, 2003, 2006, 2007, 2008, and 2012, eight years and increased in other years, reaching 51.52% and 147.18% in 2011 and 2015, respectively. Waste gas represented the largest source, accounting for 90.95% in 1992 and 96.99% in 2015, although the ratio decreased in some years.

SO2 accounted for 85.12% in 1992 to 73.74% in 2014 of the waste gas. The emissions of NOx and Soot (Dust) were 1.18 and 0.96 PJ in 2014 and they reached 3.75 and 8.93 PJ in 2015, respectively. Therefore, the growth rate of the “three wastes” in the industry increased to 147.18% in 2015. SO2 occupied most of the waste gas while NOx and Soot (Dust) grew much faster in 2015. Waste liquid containing COD and ammonia nitrogen took the third place in the first few years (1992 to 1995, 1998, 1999), then became the second one, with shares of 7.17% and 2.78% in 2014 and 2015. The fraction of ammonia nitrogen rapidly increased from 17.18% in 2001 to 31.42% in 2015 in wastewater. The solid waste decreased from 0.17 to 0.05 PJ and it was the only waste that declined during this period, implying the solid waste was controlled by the government more effectively than the waste gas and wastewater. The exergy value of “three wastes” emissions is much more than that reported in previous studies79.

The industrial “three wastes” in China showed that decreased fluctuating from 213.93 PJ in 1992–174.27 PJ in 2015. The emissions decreased in fifteen out of the twenty-four years. The fraction of waste gas increased from 48.51% (103.78 PJ), accounting for the total industrial three wastes in 1992 to 77.68% (135.37 PJ) in 2015. Among the categories of waste gases, the SO2 emission was from 64.67 to 68.50 PJ, with the maximal part from 62.32% to 50.60%; followed by soot (dust) from 39.11 to 27.71 PJ, accounting for 37.68% to 20.47%. NOx emissions decreased from 59.76 to 39.17 PJ from 2011 to 2015, accounting for 33.54% to 28.94% of the waste gas. The wastewater and waste solid declined from 97.21 to 12.94 PJ (accounting for 45.44% and 6.05% of the total industrial three wastes) in 1992 to 38.64 and 0.26PJ (22.17% and 0.15%) in 2015, respectively. “Three waste” emissions from the SPNFMI accounted for 1.56% to 12.53% of those from the industry in China over the study period. The proportion of waste gas in SPNFMI accounting for industrial waste gas enlarged from 2.92% to 15.64%. Emissions of “three wastes” from the SPNFMI increased while that in China decreased, therefore, nonferrous metals production and waste gas emissions should be supervised rigorously by the Chinese government.

The entire exergy of the environmental emissions increased from 13.01 to 138.21 PJ, of which the GHGs and “three wastes” emissions were 9.68 and 3.33 PJ in 1992 whereas they were 116.38 and 21.83 PJ in 2015, as shown in Fig. 5. The GHGs emission was much greater than the “three wastes” emission, 2.91 and 5.33 times in 1992 and 2015. Therefore, the growth rate of environmental emissions was more consistent with that of GHGs emissions. The ratio of GHGs and the ‘‘three wastes’’ was under 3 before 2004, except in 1993 and 1994, representing that the growth rate of these two emissions remained at almost the same step. The growth rate of GHGs emissions was much faster than that of the “three waste” emissions in later years.

Fig. 5
figure 5

Emissions of SPNFMI.

Full size image

Yields

Yields of nonferrous metals in China increased in the past 24 years with the yearly growth rate of 15.2%, from 46.35 to 1145.85 PJ. It was always positive except in 2009 (-0.65%) and it rebounded to 24.0% in 2010. Over the next few years, growth rate was not as fast as that in 2010 while the production still increased. Aluminium production increased 27.76 times with the fastest growth in this period, from 36.06 PJ (77.8% of the total yields) to 1037.01 PJ (90.5%). Lead and copper had an approximate development, increasing 11.08 and 11.09 times, with its yields increasing from 4.76 to 1.38 PJ in 1992 to 57.46 and 16.74 PJ in 2015, respectively. The yields of zinc and nickel expanded from 3.88 to 0.13 PJ to 33.03 and 1.02 PJ. The tin increased from 0.13 PJ in 1992–0.57 PJ in 2015, though its share in the total yields was tiny.

Indicator assessment

From 1992 to 2015, the RECC, EECC, and EERR increased 89.2%, 197.8%, and 57.4%, respectively; RLCR and ERR decreased 90.14% and 90.08% in the SPNFMI. EECC and RLCR showed the greatest changes in expansion and decline. Figure 6a shows an upward trend in RECC meaning remarkable increase in resource utilization efficiency. The rising indicator comes from the efficient utilization of energy and materials. Neverthless, it exhibited anomalies in 2012–2015, perhaps because of the overutilization of coal. The RLCR declined from 0.689 to 0.068, as shown in Fig. 6b. The expansions of natural resource as well as the slow increasing of labour and capital caused the reduction of RLCR value. The extended exergy conversion coefficient represented an increasing trend. Likely the evolutions of RECC, the EECC are lower than those of the RECC in these years (Fig. 6c). The exergetic values of EE, EM, EL, EK, ER and Yields increased 11.71, 37.50, 0.29, 0.17, 9.62 and 13.70 times, respectively. Though EM increased faster than other elements, its exergetic accounting was much less than the total extended exergy inputs. Figure 6d shows the ERR with a downward trend from 0.669 to 0.066. Due to the improvment of the industry productivity, the labour and capital were not as important as before in EEA model. Figure 6e displays the differences between EE, EM and extended exergy. The manufacturing sector in China demanded more capital and labour76, while the EE and EM of the SPNFMI increased faster than did the sum of the EL, EK and ER, leading to increase in the EEER from 0.582 to 0.916.

Through these indicators, the SPNFMI and manufacturing sector76 were compared. The indicators revealed similar trends overall and the indicators of the SPNFMI were superior to those of the manufacturing sector. From 2012 to 2020, the RECC, EECC, and EERR increased 14.2%, 26.9% and 5.6%; RLCR and ERR decreased 19.7% and 17.6% in manufacturing sector. From 2012 to 2015, the performance in these indicators differed: RECC, RLCR, EECC, and ERR decreased 24.66%, 47.6%, 20.8%, and 47.9% while EEER increased 5.2% in the SPNFMI; in manufacturing sector, RECC and EECC increased 10.4% and 14.5% while RLCR, ERR and EEER decreased 11.4%. 13.8% and 0.5%. The SPNFMI exhibits a greater dependence on EE and EM than the whole manufacturing sector. The industry demonstrates a diminishing marginal effect.

Fig. 6
figure 6

Extended exergy and exergy indicators estimation of SPNFMI during 1992–2015. (a) RECC (b) RLCR (c) EECC (d) ERR (e) EEER.

Full size image

Discussion

Consumption and production

The consumption exergy of nonferrous metals in China increased from 75.11 PJ in 1995 to 1.13 EJ in 2015, increasing 14.12 times. The growth rate of consumption was always positive, with the lowest rate of 0.81% in 2001 and the highest rate of 38.87% in 2007, showing the high consumption capacity of China. Aluminium accounted for 61.68 PJ (82.12% of total consumption) in 1995 and 1.02 EJ (90.03%) in 2015. The exergy of nickel consumption expanded 24.38 times being the fastest one.

The per capita consumption of nonferrous metals in some countries was compared in 2015. The value in China was 0.83 GJ/person, much more than that in India (0.05), higher than the United States (0.64), Japan (0.53) and Italy (0.53); lower than that in Germany (0.98) and the Republic of Korea (1.13). It was 0.06 GJ/person in 1995 increasing more than 12 times in China. Aluminium is the greatest contributor in these countries, occupying 90.03% in China (highest value) and 77.38% in the Republic of Korea (lowest value).

Both consumption and production increased from 1995 to 2015 as shown in Fig. 7. The growth rates were mostly the same step except for those in six years (2000, 2001, 2003, 2009, 2014 and 2015). The consumption was less than production most of the time except in 2000, 2009 and 2010 (three years) and the largest gap was 13.47 PJ in 2009. Aluminium dominated both consumption and production in the SPNFMI, been widely used in transportation and building for decades90,91. The growth rates of production of aluminium, the motor vehicle, floor space under construction and length of railways in operation were shown in Fig. 8. When a factor’s growth rate exceeded that of aluminum production, its relative contribution evidently increased in subsequent years. This situation occurred on four occasions: 1996, 2000, 2009 and 2011. China’s continued industrialization, especially in infrastructure, will likely sustain high aluminium consumption. Therefore, recycling aluminium might be an inevitable trend92. Widely used in semiconductor industry, Gallium is one of the byproducts in aluminium production with low yield. The production of aluminium in China far exceeded that in other countries, as did its gallium production.

Fig. 7
figure 7

Comparison of production and consumption.

Full size image
Fig. 8
figure 8

Growth rates of aluminum production and other factors.

Full size image

The top three provinces witnessed a surge in production from 12.43 to 555.40 PJ, concurrently increasing their share of the total yield from 34.47% to 53.58%. The top three provinces during the study period were drawn from Gansu, Henan, Shandong, Qinghai and Xinjiang. Shandong Province was among the top three provinces from 2004 to 2015, except in 2013. Xinjiang Province entered the top three provinces in the last three years. The aluminium production in Xinjiang increased dramatically from 0.64 PJ (accounting for 1.77% in China) in 1992 and 192.75 PJ (18.6%) in 2015, representing 300-fold increase. While production in Shandong grew by a factor of 136. In 2015, water resources in Gansu, Henan, Shandong, Qinghai and Xinjiang accounted for 0.59%, 1.03%, 0.60%, 2.11% and 3.33% of the national total, respectively. The “fresh water unit consume of aluminium” declined from 0.008 in exergy in 2007 to 0.002 in 2015. In other words, a lower water input per unit of aluminium production signified technological improvments and the production consumed valuable water resource in the water shortage provinces.

To obtain 100 units of metals, 1651.5 units of resource input comprising 947.8 units of energy, 13.4 units of material, 661.8 units of labour, 0.5 units of capital, and 28.1 units of environmental remediation were used in 1992. While in 2015, to acquire 100 units of metals, 554.6 units of resource were essential including 487.2 units of energy, 20.9 units of material, 34.5 units of labour, 0.02 units of capital, 12.1 units of environmental remediation. Therefore, to get the same quantity unit of yield, more exergy of energy was consumed than the smelting and pressing of the metals sector79 and agriculture production in Hebei province82 in China.

Energy and materials implications

The major contributors to the emissions of GHGs and the “three wastes” are coal and waste gas. Coal occupied a large part of the energy input, not as clean as other energy resources, its use could be limited strictly, such as by lifting the threshold for production in small or new coal mines and importing more natural gas or petroleum to upgrade the energy structure of the SPNFMI. For considerable electricity deriving from thermal power in China, large electricity consumption inevitably leads to emissions indirectly. Hence China may aim to develop clean energy, especially renewable energy for the sustainable development of the industry.

Substantial resource depletion led stress to China. The ratio of emissions to material inputs decreased from 2.10 in 1992 to 0.58 in 2015. Bauxite, alumina and copper ore are the key point leading expansion of dependency on material resources abroad. First, China could encourage communication with countries which export high exergy content ores, such as Australia, Brazil, Chile and Congo or large quantities of scrap metals; furthermore, diversifying its import sources would enhance supply chain resilience. Second, developing scrap nonferrous metals domestically to lower emissions could not be implemented quickly for inadequate scrap accumulation93, while this is the developing direction in future. Therefore, the shortage of nonferrous metals resources could be alleviated to some extent.

Seize the “The Belt and Road” opportunity and promote the international cooperation of the industry. “The Belt and Road” project covers several countries or regions that are undergoing industrialization and urbanization, causing substantial demand for nonferrous metals. China may catch the chance and elevate international cooperation to digest the excess production. “The Belt and Road” project includes countries with mineral resources of nonferrous metals. Moreover, the “dual-circulation” development paradigm (with the big domestic cycle as the main body, the domestic and international double cycles promote each other) was proposed to foster the sustainable development of the SPNFMI.

Based on the thermodynamics law, EEA unifies the natural resources, labour, capital and emissions in joules from the sustainable usefulness point78. EEA is a tool to measure different resources consumption identifying the environmental impact and compare the resources utilization and structure in different industries or regions94. As the product of the SPNFMI, metals have dissimilar characteristics. EEA theory reflects the degradation of metals from the thermodynamics point, ignoring the physical, chemical and other properties of nonferrous metals. Different with other researches accounting the labour through monetary approach, EEA delineates extended exergetic content of labour in production activities by the transactions of commodities, services and labour17. Capital does not only comprise the labour wages in the real world. This theory does not consider such scenes as credit or monetary overspending for the labour and capital leading to uncertainties of capital movement. The exergoeconomics in EEA is a cost allocation process paralleling the supply of economics, not obeying the second thermodynamic law strictly94. The calculation of the ER fails to reveal the toxicity and negative impact of kinds of emissions on the environment in different regions and long term95. In the future, quantifying the labour and environmental influence more reasonably may be a valuable work. Provinces of China had different features in nonferrous metals production and the characteristic may be explored in EEA research. With the estimation of EEA, conclusions from biophysical sustainability judgment, economic, social and environmental development methods can be revised for nations, regions and industries in future.

Conclusions

Given its large population, upgrading life level and large nonferrous metal demand, China’s SPNFMI should improve resource utilization efficiency to achieve sustainable development. From a thermodynamic perspective, energy dominated the total input resources. Therefore promoting the sustainable development of the energy-intensive industry necessitates a dual strategy: reducing fossil energy consumption and encouraging renewable energy consumption . Given the low exergy content of domestic resources, material supplies should be allocated on global scale. For aluminum’s extensive applications, its production in China may continue to grow despite surpassing output levels of other countries.