Uncovering Cleaner Method for Underground Metal Mining: Enterprise-Level Assessment for Current and Future Energy Consumption and Carbon Emission from Life-Cycle Perspective

: China has committed to peak its carbon emissions by 2030, which puts forward a new issue for underground metal mines—selecting a cleaner mining method which requires less energy and generates less carbon emissions. This paper proposes an enterprise-level model to estimate life-cycle energy consumption and carbon emissions, which takes more carbon sources (e.g., cement and carbon sink loss) into consideration to provide more comprehensive insights. Moreover, this model is integrated with the energy-conservation supply curve and the carbon abatement cost curve to involve production capacity utilization in the prediction of future performance. These two approaches are applied to 30 underground iron mines. The results show that (1) caving-based cases have lower energy consumption and carbon emissions, i.e., 673.64 GJ/kt ore, 52.21 GJ/kt ore (only considering electricity and fossil fuel), and 12.11 CO 2 eq/kt ore, as compared the backﬁlling-based cases, i.e., 710.08 GJ/kt ore, 63.70 GJ/kt ore, and 40.50 t CO 2 eq/kt ore; (2) caving-based cases present higher carbon-abatement potential (more than 12.95%) than the backﬁlling-based vases (less than 9.68%); (3) improving capacity utilization facilitates unit cost reduction to mitigate energy consumption and carbon emissions, and the energy-conservation and carbon-abatement potentials will be developed accordingly.


Introduction
Under the driving effect of more than 1.4 billion people and fast-growing economy [1], China has been the largest carbon emitter since 2007, as well as the largest energy consumer since 2010 [2,3]. To mitigate global climate change, China announced a series of ambitious long-term targets, e.g., peaking carbon emission by 2030 [4], and carbon neutrality around 2060 [5]. This poses a new problem for China's underground metal mines-selecting a cleaner method to extract minerals, which requires less energy and generates fewer carbon emissions [6].
Open-stope, caving, and backfilling are three conventional and mainstream mining methods for underground metal mines [7][8][9][10]. The caving method is favored due to high cost efficiency [11], but the surface subsidence associated still heavily restricts its application [12,13]. Considering the increasing public concern for surface subsidence [14], more and more projects employ the backfilling method, because it can preserve the ground surface by backfilling and supporting the void after deposit excavation [15]. Moreover, many projects initially by open stope also backfill their voids subsequently, to protect the ground Therefore, due to the absence of either a detailed inventory or robust approach, it is still an intractable issue to uncover the cleaner mining method from open-stope, caving, and backfilling, especially under the consideration of both current and future performance regarding energy consumption and carbon emissions of metal miners from a life-cycle perspective. In order to bridge this research gap, we established an enterprise-level model to estimate present life-cycle energy consumption and carbon emissions. Moreover, this lifecycle model is integrated with the CSC and ACC approach to predict future performance. Eventually, 30 underground iron mines are involved as case studies to uncover the cleanest mining method. This work contributes to the existing literature from both theoretical and practical perspectives. Theoretically, it provides a robust approach to test the current and future life-cycle energy consumption and carbon emissions of metal mineral extraction by underground mining to guide policy design or improvement. In practice, this model can provide detailed predictions of various technologies to mitigate energy consumption and carbon emissions in different projects, which means it facilitates the determination of whether a technology should be employed.

System Boundary Definition
The research scope of this study is the entire life cycle of minerals extraction in underground metal mines, from deposit discovery to raw ore being transported to dressing plants. This life cycle can be categorized into nine stages, e.g., ventilation, drilling, blasting, supporting, loading, haulage, crushing, backfilling, and dewatering. Three type of energy are considered in the life-cycle assessment for energy consumption, i.e., electricity, fossil fuels, and industrial explosives. In addition to these energies, we also take cement and carbon sink loss into account as carbon sources, which are rarely assessed as such in the existing literature. This system boundary defined is valid not only for assessing lifecycle energy consumption and carbon emissions at present, but also for predicting energy conservation and carbon abatement potential in the future. Additionally, it should be noted that the functional unit has been defined to be per kiloton extracted metal ore to unify the results of various indicators from different cases.

Life Cycle Assessment for Energy Consumptions and Carbon Emissions
The model to estimate life-cycle energy consumptions and carbon emissions by energy type (or carbon source) is described below.

Electricity
Electricity is the predominant energy source in underground metal mines to maintain a favorable environment at the underground working faces [40]. Equations (1) and (2) are proposed to quantify the energy consumption and indirect carbon emissions from electricity-powered equipment, respectively.
where EC EL and CE EL are the unit energy consumption and unit carbon emission from electricity-powered mining equipment, respectively; α is a coefficient to standardize the unit of energy consumption; EF EL is the carbon emission factor of electricity; P i is the power of mining equipment i; t EL i is the operating time of electricity-powered mining equipment i annually; and Q Ore is the annual yield of minerals in a underground metal mine.

Fossil Fuel
Fossil fuel is another primary energy source, especially in the development engineering, due to its advantage of flexibility. Consuming fossil fuel leads to both direct and indirect carbon emissions which occur in their production and combustion, respectively [41][42][43]. The energy consumption and carbon emissions related to fossil fuels can be calculated by Equations (3) and (4), respectively.
where EC FF and CE FF are the unit energy consumption and unit carbon emissions related to fossil fuels, respectively; NCV j is the net calorific value of the fossil fuel j; EF FF j is s the carbon emission factor of fossil fuel j; Q i,j Q FF i,j is the consumption of fossil fuel j by mining equipment i in unit time; and t FF i is the annual operating time of mining equipment i powered by fossil fuel.

Industrial Explosive
Underground metal mines utilize the energy released by industrial explosives in blasting reactions to break intact hard rocks. Producing industrial explosives generates indirect carbon emissions, but the direct emissions depend on the proportion of blasting agents and additives [44]. Equations (5) and (6) are valid to estimate the energy released in blasting reaction and the carbon emissions, respectively.
where EC IE and CE IE are the unit energy consumption and unit carbon emission due to industrial explosives, respectively; DH j is the detonation heat of the industrial explosive j; EF IE j is the carbon emission factor of the industrial explosive j; and Q IE j is the consumption of industrial explosive j to break unit quantity of rocks.

Cement
Cement is required in underground metal mines using either the open-stope, caving or backfilling methods to enhance the stability of shafts, tunnels, and chambers. But backfilling-based mines require much more cement to cement the materials backfilled into the void after deposit excavation [45]. Although no energy consumption related to cement takes place in the mining site, indirect carbon emissions still occur, such as carbon dioxide emitted in limestone quarries and clinker plants [46,47]. Such indirect carbon emissions can be quantitively described by Equation (7).
where CE C is the unit carbon emission due to consumed cement; EF C is the carbon emission factor of cement; and Q C is the annual consumption of cement.

Carbon Sink Loss
In the underground metal mines using the open stope or caving methods, large-scale subsidence is commonly observed on ground surface [48]. The destruction of vegetation raises direct carbon emissions, which can be quantitatively described by the net primary production (NPP) [49,50]. And the carbon emission can be calculated by Equation (8), accordingly.
where CE CSL is the unit carbon emission due to carbon sink loss; β is a coefficient to convert the NPP to carbon emissions; NPP 0 and NPP are the NPPs of the vegetation on the ground surface before and after subsidence, respectively; and S is the area of the pit formed after surface subsidence. If the inputs listed in Equations (1)- (8) have been provided, the life-cycle unit energy consumption and unit carbon emission in an underground metal mine can be estimated by Equations (9) and (10), respectively.
where EC and CE are the unit energy consumption and unit carbon emission of the minerals in the life cycle from deposit to the gate of dressing plants, respectively; and EC k and CE k are the unit energy consumption and unit carbon emission due to the energy or carbon source k.
Equations (1)-(10) are the model proposed to assess the life-cycle energy consumption and carbon emission of the minerals extracted from underground metal mines.
Compared with existing models, this approach extends the scope of energy and carbon sources by taking industrial explosive, cement, and carbon sink loss into consideration. Hence, it can be expected to provide comprehensive insight into the environmental impact of underground metal mining sector. Moreover, this approach utilizes engineering-based data as inputs, this enables it to relate energy consumption and carbon emission directly to the mining equipment and stage. This means it has the potential to be integrated with the CSC and ACC approach to include production capacity utilization in assessments of energy-conservation and carbon-abatement potentials.

Bottom-Up Assessment for Energy-Conservation and Carbon-Abatement Potential
Mining method selection is a strategic decision, which means not only the current but also the future performance of energy consumption and carbon emissions are critical concerns. To develop the accuracy of the assessment for energy-conservation and carbon-abatement potentials on enterprise level, we integrate the LCA model proposed with the ECSC and CACC approaches to take capacity utilization into account. The contribution of energy-saving technology and reduced carbon emissions can be estimated by Equations (11) and (12), respectively.
where ECM n,t and CEM n,t are the potentials of technology n to mitigate energy consumption and carbon emission in year t, respectively; EC i t and CE i t are the energy consumption and carbon emissions due to mining equipment i in year t occurring on-site, which can be estimated by the LCA model proposed; EC CAP and CE CAP are the energy consumption and carbon emissions in the original design, respectively; UECM 0 n and UCEM 0 n are unit contribution of technology n to saving energy and reducing carbon emissions, which can be derived from the illustrative examples employing this technology; and CAP i n is the installed capacity of mining equipment i, to which that technology n will be deployed.
On the other hand, the investment in a technology still depends on the installed capacity, and the associated cost to save unit energy or reduce unit carbon emissions can be calculated by Equations (13) and (14), respectively.
where C ECM n and C CEM n are the unit cost to mitigate energy consumption and carbon emission by employing technology n; t n is the lifetime of technology n; γ is a coefficient to spread investment for technology n over its lifetime, which can be calculated by r is the discount rate; U I n is the unit investment for technology n, which can be calculated derived from the illustrative examples; ∆OM n,t is the additional cost for operating and maintaining technology n; γ is a coefficient to spread investment for technology n over its lifetime, which can be calculated by γ = r· 1 − (1 + r) −t n ; r is the discount rate; and P C and P E are the price of carbon and energy, respectively.
Equations (13) and (14) enable the break-even analysis to determine the employment of a technology [51], such as Equations (15) and (16).
where ∆ E n and ∆ C n are the indicators for the cost-effectiveness of technology n; and the employment of a technology can be determined by testing whether ∆ E n > 0 or ∆ C n > 0 is satisfied.
Eventually, the potential of an underground metal mine to save energy or reduce carbon emissions can be predicted by accumulating the contributions of these cost-effective technologies. Compared with the conventional ECSC and CACC approaches, this approach eliminates the negative impact of some assumptions on the results. Firstly, the production capacity utilization is considered in accordance with the difference between on-site energy consumption (or carbon emissions) and the value in original design. Secondly, the assumption of constant unit cost is avoided.

Data Source
In total, 30 of China's underground iron mines from 11 provinces were selected for this study, including 17 caving-based and 13 backfilling-based cases, as illustrated in Figure 1. The total yield is more than 70 million t iron ore per year, comprising~34% of China's iron ore production from underground projects in 2019. The results proposed in this work can be expected to represent the national common level. An inventory analysis was conducted to obtain the inputs for estimating life-cycle of energy consumptions and carbon emissions, as presented in Appendix A.
We applied the developed ECSC and CACC approaches to 8 projects from these 30 cases to uncover the difference between caving-based or backfilling-based cases to save energy and reduce carbon emissions. Table 1 lists 15 the energy conservation and carbon abatement (ECCA) technologies, promoted to the underground mining sector by China's NDRC [52]. More detailed information is provided in Appendix B. Additionally, some assumptions should be noted, i.e., the discount rate is 15% [53], and the energy and carbon prices equal to 100 CNY/GJ and 100 CNY/t CO 2 eq, respectively. We applied the developed ECSC and CACC approaches to 8 projects from these 30 cases to uncover the difference between caving-based or backfilling-based cases to save energy and reduce carbon emissions. Table 1 lists 15 the energy conservation and carbon abatement (ECCA) technologies, promoted to the underground mining sector by China's NDRC [52]. More detailed information is provided in Appendix B. Additionally, some assumptions should be noted, i.e., the discount rate is 15% [53], and the energy and carbon prices equal to 100 CNY/GJ and 100 CNY/t CO2 eq, respectively.    and 710.08 GJ/kt, respectively. However, such results are still insufficient to demonstrate that the caving method is more effective in energy consumption than the backfilling, due to the predominant role of explosive-related energy. For instance, the energy consumption in Jin'an (B09, 605.83 GJ/kt) is higher than that in Xiaoguanzhuang (C09, 595.73 GJ/kt). But if the explosive-related energy is ruled out, Jin'an's energy consumption (36.14 GJ/kt) falls below that in Xiaoguanzhuang (64.97 GJ/kt). Thus, we compared only the energy consumption involving electricity and diesel between the caving-based and backfillingbased cases, to eliminate the interference of industrial explosives, as listed in Table 2.

Energy Consumptions and Carbon Emissions
3.1.1. Life-Cycle Energy Consumption Figure 2 provides the results of life cycle assessment for energy consumption in 30 underground iron mines. We calculated the average energy consumption in either cavingbased or backfilling-based cases to test the difference, and the results were 673.64 and 710.08 GJ/kt, respectively. However, such results are still insufficient to demonstrate that the caving method is more effective in energy consumption than the backfilling, due to the predominant role of explosive-related energy. For instance, the energy consumption in Jin'an (B09, 605.83 GJ/kt) is higher than that in Xiaoguanzhuang (C09, 595.73 GJ/kt). But if the explosive-related energy is ruled out, Jin'an's energy consumption (36.14 GJ/kt) falls below that in Xiaoguanzhuang (64.97 GJ/kt). Thus, we compared only the energy consumption involving electricity and diesel between the caving-based and backfilling-based cases, to eliminate the interference of industrial explosives, as listed in Table 2.  When only the energy consumption related to electricity and fossil fuel are considered, Table 2 shows more that caving-based cases (9 of 17) present lower energy consumption (colored by blue), than backfilling-based cases (5 of 13). Particularly, the energy consumption due to electricity was much lower in the caving-based (38.85 GJ/kt, average-weighted), than in the cases using the backfilling method (56.68 GJ/kt, average-weighted). This accounts for the phenomenon that caving-based cases require less energy (52.21 GJ/kt, average-weighted) than the backfilling-based (63.70 GJ/kt, average-weighted), although they consume more fossil fuels (13.36 GJ/kt and 7.02 GJ/kt for the caving-based and backfilling-based average-weighted, respectively). Thus, the results of the LCA illustrated in Figure 2 and Table 2 demonstrate that caving method is superior to backfilling in energy consumption. Figure 3 provides the carbon intensity supply curve of the 30 underground iron mines, and it explicitly shows that the carbon emissions in caving-based cases are much lower than that in the backfilling-based cases. In the caving-based cases, the carbon emissions varied between 4.73 t CO 2 eq/kt (C02, Gongchangling) and 20.25 t CO 2 eq/kt (C15, Jinshandian), and the weighted average was 12.11 t CO 2 eq/kt. Meanwhile, these values in the backfillingbased increases from 32.30 t CO 2 eq/kt (B07, Wuji) to 50.47 t CO 2 eq/kt (B05, Jinling), and the weighted average was 40.50 t CO 2 eq/kt. This means that the caving method is also superior to backfilling in carbon emissions.  The contribution of each carbon source is also illustrated in Figure 3, which explains the underlying drivers for the difference of carbon emission in either the caving-based or backfilling-based cases. It can be observed that the carbon emissions in the caving-based cases were dominated by electricity (more than 68.32%), and the contribution of carbon sink loss was less than 5.67%. In some caving-based cases, such as C01 (Banshi), C02 (Gongchangling), C04 (Xiaowanggou), C05 (Maogong), C06 (Heishan), C07 (Xingshan), The contribution of each carbon source is also illustrated in Figure 3, which explains the underlying drivers for the difference of carbon emission in either the caving-based or backfilling-based cases. It can be observed that the carbon emissions in the cavingbased cases were dominated by electricity (more than 68.32%), and the contribution of carbon sink loss was less than 5.67%. In some caving-based cases, such as C01 (Banshi), C02 (Gongchangling), C04 (Xiaowanggou), C05 (Maogong), C06 (Heishan), C07 (Xingshan), C11 (Jingtieshan), C16 (Jining), and C17 (Shilu), no carbon emissions related to carbon sink loss occurred, because the underground mining was implemented after surface mining. Surface mining, rather than underground caving, is responsible for the destruction of ground vegetation and the associated carbon sink loss.

Life-Cycle Carbon Emissions
In the backfilling-based cases, cement is another primary contributor to carbon emission, besides electricity. And cement's contribution heavily depends on the cement-sand ratio of the slurry backfilled into the void after deposit excavation. For instance, in cases utilizing a lower cement-sand ratio slurry, e.g., B03 (Fuquan, cement-sand ratio is 1:12) and B10 (Longtangyan, cement-sand ratio is 1:10), the contribution is obviously lower than that in other cases (cement-sand ratio higher than 1:8). This reveals that lowering the cement-sand ratio is a robust approach to mitigate carbon emissions in backfillingbased underground metal mines, in addition to the conventional approaches focused on conserving electricity and fossil fuel.
It should be noted that the energy consumption in China's caving-based iron mines ( Table 2) is similar to the inventory in the existing literature. For instance, the energy consumption in Kiruna iron mine, Sweden, is 46.728 GJ/kt, and contribution of electricity and fossil fuel are 41.4 and 5.328 GJ/kt, respectively [54]. Such results validate the results in this study, although it is difficult to compare the results of carbon emissions directly because the carbon emission factor of electricity significantly varies in different countries. The results of life cycle assessments for energy consumption and carbon emissions in 30 underground metal mines demonstrate that caving is cleaner than backfilling for deposit excavation. This phenomenon is opposite to the view in the existing literature [18,19], because those studies only considered environmental impacts due to surface destruction, rather than from a life cycle perspective. This indicates a one-sided preference is very likely to mislead policy decisions. For instance, the promotion of the backfilling method will raise the energy consumption and carbon emissions of China's metal mining sector. On the other hand, the results also indicate that energy conservation and carbon abatement actions for underground metal mines should primarily focus on mitigating electricity consumption, but additional attention is required to reducing the cement-sand ratio of backfilling slurry in backfilling-based mines. Figure 4 shows the ECSCs of 15 ECCA technologies (Table 2), when they are applied to the selected caving-based (i.e., Xiaowanggou (C03), Yanqianshan (C04), Maogong (C05), and Shilu (C17)), and backfilling-based cases (i.e., Luohe (B06), Wuji (B07), Lilou (B08), and Longtangyan (B10)). The production capacity utilization was calculated based on the energy consumption and carbon emissions occurring in 2019.

Energy Conservation Potentials
As can be seen from Figure 4, the unit costs to save energy present significant variations in the different cases. Yanqianshan (Figure 4b) present a much higher unit cost than other cases, because Yanqianshan's capacity utilization is lower than all the other cases, i.e., production capacity is 8 million t/a, and the outputs in 2019 is 2.9 million 8 million t/a. Such a low capacity utilization raises the unit cost to conserve energy, and further reduces the potential of Yanqianshan to save energy. When the prices of energy and carbon are equal to 120 CNY/GJ and 100 CNY/t CO 2 eq, Yanqianshan has the potential to reduce its energy consumption from 45.89 GJ/kt (without the consideration for the energy released in blasting reactions by industrial explosives, as listed in Table 2 The vertical axis is the unit cost for saving energy, in k CNY/GJ, and the horizontal axis is the cumulative energy conservation potential, in TJ/a. The dotted line in red represents the break-even price of energy (PE), which equals 120 CNY/GJ. The colors indicate the type of energy that the ECCA technology targets for conservation, i.e., green and yellow represent electricity and fossil fuel, respectively. The vertical axis is the unit cost for saving energy, in k CNY/GJ, and the horizontal axis is the cumulative energy conservation potential, in TJ/a. The dotted line in red represents the break-even price of energy (PE), which equals 120 CNY/GJ. The colors indicate the type of energy that the ECCA technology targets for conservation, i.e., green and yellow represent electricity and fossil fuel, respectively.
Apart from the case of Yanqianshan, the energy conservation potentials of the involved cases all varied around 40%, although the ECSCs present significant variation in each case. This reveals no significant difference of energy conservation potential can be observed between the caving-based and backfilling-based cases. However, because backfilling-based cases requires more energy to extract unit minerals at present, it can be expected the higher energy consumption in the projects using the backfilling method will continue in the future. Figure 5 provides the CACCs of 15 ECCA technologies, when they were deployed to the caving-based (i.e., Xiaowanggou (C03), Yanqianshan (C04), Maogong (C05), and Shilu (C17), or backfilling-based cases (i.e., Luohe (B06), Wuji (B07), Lilou (B08), and Longtangyan (B10). Because the benefit from conserved energy surpasses the investment and additional operating and maintainance cost for employing these technologies, the unit carbon abatement cost of some technologies presents a negative value, such as T01 in Xiaowanggou (−1063.33 CNY/t CO 2 eq), T12 in Yanqianshan (−468.12 CNY/t CO 2 eq), T03 in Luohe (−1119.57 CNY/t CO 2 eq), and T04 in Wuji (−994.98 CNY/t CO 2 eq). This means these technologies will be utilized, no matter how high the carbon price.

Carbon Abatement Potentials
The CAACs illustrated in Figure 5 enable the prediction of the future performance of carbon emissions in underground iron mines. When the prices of energy and carbon are equal to 120 CNY/GJ and 100 CNY/t CO 2 eq, the caving-based cases have the potentials to reduce the carbon emissions by 12.95% (in Xiaowanggou, from 14.38 to 12.52 t CO 2 eq/kt), 5.20% (in Yanqianshan, from 12.55 to 11.89 t CO 2 eq/kt), 13.73% (in Maogong, from 12.26 to 10.58 t CO 2 eq/kt), and 15.70% (in Shilu, from 13.81 to 11.64 t CO 2 eq/kt), respectively. The lower production utilization in Yanqianshan is responsible for this difference, because it raises the unit cost to reduce carbon emissions. Consequently, if Yanqianshan's production were to reach its designed capacity, its carbon abatement potential will significantly increase, because more ECCA technologies will be effective in the break-even analysis.
On the other hand, backfilling-based cases presented significant lower carbon-abatement potential than the caving-based cases, except Yanqianshan. For instance, the cases of Luohe, Wuji, Lilou, and Longtangyan had the potential to reduce their carbon emissions by 7.02% (from 42.51 to 39.62 t CO 2 eq/kt), 2.62% (from 32.30 to 31.45 t CO 2 eq/kt), 5.23% (from 40.56 to 38.44 t CO 2 eq/kt), and 9.68% (from 32.90 to 29.72 t CO 2 eq/kt), respectively. The absence of technology to lower the cement-sand ratio in backfilling slurry accounts for this phenomenon. As depicted in Table 1, ECCA technologies in the mining sector primarily work by saving electricity and fossil fuels, but the predominant role of cement in the carbon emissions from caving-based mining eliminates their potential to reduce carbon emissions. This demonstrates that technologies which can save cement and cement consumption in the backfilling stage are necessary for backfilling-based underground metal mines, in addition to the conventional ones promoted by China's NDRC.  )). The vertical axis is the unit cost for mitigating carbon emissions, in k CNY/t CO 2 eq, and the horizontal axis is the cumulative carbon abatement potential in kt CO 2 eq/a. The dotted line in red represents the break-even price of carbon emissions (PC), which equals to 100 CNY/t CO 2 eq. The colors indicate the type of carbon source that the ECCA technology targets to save, e.g., green and yellow represent electricity and fossil fuel, respectively.

Policy Implications
From the analysis in this work, we propose the following suggestions for policy design and implementation.
Firstly, before the policy-making process, a standardized model is necessary to assess the environmental impact of the underground metal mining sector, comprehensively. Existing studies regard backfilling as the cleaner mining method [18,19], because it can protect the ground surface from large-scale destruction. However, the quantitative results from a life-cycle perspective (Figures 2-5) explicitly show caving is cleaner. This phenomenon reveals that this one-sided preference is very likely to mislead the policy decisions, especially in the absence of sufficient inventory. Thus, policymakers need a series of standardized models to assess the various impacts of underground metal mines on the environment, which should emphasize indicator selection, system boundary definition, and the preference of policies.
Secondly, Chinese governments can promote the caving method for peaking the carbon emissions from underground metal mining sector by 2030. The results from 30 underground iron mines (Figures 3 and 4) show the caving-based mining has lower unit carbon emissions and higher carbon-abatement potential. If the productivity by the backfilling method is replaced by the caving method, the overall carbon emissions of China's metal supply will decrease accordingly.
Thirdly, China's government can implement a mandatory policy to define a threshold of production capacity utilization for underground metal mining sector. The results in Section 3.2 explicitly demonstrate that developing production capacity utilization facilitates a reduction to the unit cost of ECCA technologies to save energy and reduce carbon emissions. This means the threshold of production capacity utilization is an indirect but robust approach to motivate the enthusiasm of underground metal mines, because employing cost-effective ECCA technologies is a profitable decision.

Conclusions
To peak the carbon emissions in underground mining sector by 2030, uncovering a cleaner method has been a critical concern for both mining enterprise and China's government. This paper proposes an enterprise-level model under the LCA framework to estimate the energy consumption and carbon emissions of underground metal mines, integrated with CSC and ACC, to include production capacity utilization in predictions of future performance. Detailed data from 30 underground iron mines are provided, which is a contribution due to the absence of sufficient inventory in the existing literature.
The results of the LCA demonstrate that caving is the cleaner method for deposit excavation from a life-cycle perspective. These cases require less energy (673.64 GJ/kt, 52.21 GJ/kt without industrial explosive) and generate less carbon emissions (12.11 t CO 2 eq/kt), than backfilling-based mines, i.e., 710.08 GJ/kt, 56.68 GJ/kt without industrial explosive, and 40.50 t CO 2 eq/kt. Additionally, although no significant difference of energy conservation potential can be seen among the cases using either the caving or backfilling methods, the carbon abatement potential in the caving-based method is much higher (from 5.20% to 15.70%) than in the backfilling-based cases (from 2.62% to 9.68%). These results show that backfilling is the cleaner mining method, and that the existing literature is no longer valid from life-cycle perspective, because they demonstrate a one-sided preference very likely to mislead policy decisions.
In accordance with the engineering-based data from 30 underground iron mines, some implications for policy design can be summarized. Firstly, a comprehensive inventory and standardized model, especially from a life-cycle perspective, is suggested to be employed for policy decisions, to mitigate the effect of the one-sided preference. Secondly, to peak the carbon emissions from China's metal mining sector, the promotion of the caving method is suggested. Thirdly, a mandatory policy to improve production capacity utilization is also proposed, because this tactic facilitates energy savings and reduces carbon emissions of underground metal mines by reducing the unit cost for employing ECCA technologies.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to request by the involved cases.

Conflicts of Interest:
The authors declare no conflict of interest. Table A1. Heat values and carbon emission factors of energy or carbon sources.