Environmental and mine closure costs in the block sequencing of a coal deposit

Original article Environmental and mine closure costs in the block sequencing of a coal deposit David Oliveros-Sepúlveda a,* , Marc Bascompta-Massan´ es a,1 , Giovanni Franco-Sepúlveda b,1 a Universitat polit` ecnica de Catalunya, Escola Polit` ecnica Superior d’Enginyeria de Manresa. Manresa, Edifici MN1. Av. de les Bases de Manresa, 61-73 08242 Manresa, Spain b Universidad Nacional de Colombia, Facultad de Minas. Av. 80 #65 - 223, Villa Flora, Medellín, Robledo, Medellín, Antioquia, Colombia ARTICLE INFO Keywords: Strategic mine planning Environmental costs Closure costs Social costs ABSTRACT At present, extractive industries face significant sustainability challenges in the territories where they operate. Traditionally, environmental and social variables have not been integrated into the processes of optimization and sequencing in mining plans, leading to outcomes that are not holistic and are accompanied by higher risks. Efforts to reduce geological, economic, and operational uncertainty have seen considerable progress in recent years. However, limited research addresses the environmental and social uncertainties that mining projects may face. This research proposes the internalization of the costs associated with the closure of the mine, environmental costs and costs of conflict resolution, calculated based on equations suggested by various authors in the state of the art. The base case of an open pit coal mine located in Colombia with previously established initial conditions and associated costs, calculated after the optimization process, is addressed. Subsequently, a new case is generated in which these costs are charged to the economic valuation of the mineral, punishing its calorific value. At the end, a sensitivity analysis is carried out with the future projections of the price of coal. The results obtained indicate that, by internalizing these costs, it is possible to reduce the investment risk and there is an improvement in the operational, social and environmental performance of the case study. Results that could benefit the company, the State and the communities. 1. Introduction In recent decades, the integration of environmental variables into mine planning has gained increasing significance, driven by the growing regulatory demands and the need to adopt more sustainable mining practices. In certain regions, particularly in developing countries, mining operations have left a substantial environmental legacy, affecting local ecosystems and communities. As a result, evaluating and effectively managing these impacts has become a critical challenge for the industry, especially with the advent of sustainable development concepts and the Sustainable Development Goals (SDGs) (UN, 2015). Traditionally, resource extraction optimization focused on economic and operational parameters. However, there has been a growing recognition that environmental and social factors must also be considered, as failure to do so may result in additional costs for both companies and governments (Lechner et al., 2016; McCullough et al., 2018). Several studies suggest that mine planning should be approached holistically, aiming not only to maximize economic benefits but also to minimize negative environmental impacts (Heydari and Osanloo, 2024; Liu et al., 2024). The costs associated with rehabilitating affected areas and mitigating unforeseen environmental consequences are often significantly higher when addressed at the end of the mine’s life cycle, rather than being progressively integrated into the early planning stages (Getty and Morrison-Saunders, 2020). Conducting ex-ante assessments of environmental impacts and incorporating them into block sequencing models and final pit designs are crucial for ensuring long-term sustainability (Kosinskiy et al., 2019; X. Xu et al., 2017). In this context, the legislation in countries such as Chile, the United States, and Australia mandates that mining companies account for these costs in their evaluations, employing clear methodologies for * Correspondence author at. Universitat polit` ecnica de Catalunya, Escola Polit` ecnica Superior d’Enginyeria de Manresa. Manresa, Edifici MN1. Av. de les Bases de Manresa, 61-73 08242 Manresa, Spain. E-mail addresses: [email protected] (D. Oliveros-Sepúlveda), [email protected] (M. Bascompta-Massan´ es), [email protected] (G. FrancoSepúlveda). 1 Principal author Contents lists available at ScienceDirect The Extractive Industries and Society journal homepage: www.elsevier.com/locate/exis https://doi.org/10.1016/j.exis.2025.101668 Received 5 February 2025; Received in revised form 3 April 2025; Accepted 5 April 2025 The Extractive Industries and Society 23 (2025) 101668 Available online 18 April 2025 2214-790X/© 2025 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/bync/4.0/ ). quantification. This represents a significant step toward embedding sustainability into mining projects. However, despite regulatory progress, notable gaps remain in the methodology for accurately quantifying environmental and closure costs (Oblasser and Chaparro, 2008; Pietrzyk-Sokulska et al., 2015). As a result, recent studies propose various valuation methodologies that encompass not only direct rehabilitation costs but also those arising from the loss of ecosystem services and carbon emissions (Rahmanpour and Osanloo, 2017; X. C. Xu et al., 2014). A key approach to addressing these challenges is the application of optimization techniques that integrate environmental variables into the mine planning process. Various algorithms and simulation models, such as those based on Monte Carlo simulations and decision tree analysis, have been proposed to estimate the environmental costs linked to mining extraction and rehabilitation (Paricheh and Osanloo, 2017). These approaches enable not only the quantification of both direct and indirect impacts but also the evaluation of alternative scenarios to identify the most effective exploitation and closure strategies (Rahimi and Ghasemzadeh, 2015). Moreover, recent studies underscore the importance of incorporating techniques like the Preference Ranking Organization Method for Enrichment Evaluation (TOPSIS) and grey clustering, among others, in multicriteria decision analysis for mine planning (Adibi et al., 2015). These methods facilitate the integration of technical, economic, social, and environmental factors, empowering more informed and responsible decision-making (Moradi and Osanloo, 2015; Tabesh et al., 2023). While significant research exists on integrating environmental variables into mine planning, few studies directly incorporate environmental and closure costs into block models comprehensively. In terms of closure costs, recent research by Xu et al. (X. Xu et al., 2017) and Nehring and Cheng (Nehring and Cheng, 2016) has demonstrated that extending a project’s lifespan by incorporating closure costs early in the planning process can yield economic benefits beyond initial projections. The early inclusion of these costs and the scheduling of progressive closures throughout the project life cycle are more cost-effective than addressing them reactively at the end of the exploitation phase (X. Chuan Xu et al., 2018). Progressive closure methodologies, supported by Geographic Information Systems (GIS) and other analytical tools, have proven effective in monitoring environmental impacts and adjusting mining operations to prevent or mitigate further damage (Liao et al., 2013). Furthermore, the economic valuation of ecosystem services, both direct and indirect, is increasingly recognized as an integral component of mining project evaluation (X. Xu et al., 2017). These costs, which encompass biodiversity loss, diminished air and water quality, and negative effects on soils and landscapes, must be incorporated into optimization models to more accurately reflect the true costs of mining operations (Gu et al., 2013). In this regard, research by Kosinskiy et al. (2019) and Rahmanpour and Osanloo (2017) emphasizes the importance of quantifying the ecological footprint and the damage inflicted on ecosystem functions by extractive activities, which is critical for adopting a sustainable mining approach. To reduce risks and provide alternative economic evaluations for both companies and governments, this study focuses on the inclusion of environmental, social, and mine closure costs for an open-pit coal deposit in northern Colombia, quantified using methodologies identified in the literature. Building on previous research, an approach will be developed to optimize mine planning by integrating these costs early in the operation, incorporating them into block sequencing and final pit selection, to achieve efficient and sustainable extraction. 2. Materials and methods After identifying the state of the art and the conceptual framework present in the literature, the key aspects of the case study were addressed, including its technical, environmental, and social characteristics, alongside the corresponding economic valuation. Given the morphology of the mineral deposit and the operational conditions of the project, it was determined that the previously calculated costs should be integrated into the geological model. This approach is intended to be applied specifically to each deposit, as the unique characteristics of each case study in the literature preclude the direct application of results to other projects. To quantify the costs associated with both direct and indirect ecosystem services (environmental costs), the methodology proposed by Xiao-Chuan et al. (X. C. Xu et al., 2014) was selected, as it provides clear and concise guidelines for obtaining a comprehensive economic evaluation in mining projects. For calculating the costs related to mine closure activities, reference was made to Law 20,551 of SERNAGEOMIN (National Service of Geology and Mining) (Sernageomin, 2014). However, due to differences in natural conditions compared to the case study, some modifications and correction factors were applied. For a more detailed understanding of the equations and methodologies employed, readers are encouraged to consult the original sources directly. The costs associated with conflict resolution were estimated based on calculations referring to the number of people directly affected, the average number of individuals per household, and a hypothetical economic compensation amount that individuals would be willing to accept annually for the impacts generated by the mine. Based on the geological model, and utilizing Deswik mine planning software, two mining schedules and block sequencing scenarios were developed. The first corresponds to "Case A," where traditional planning is applied, and the second, "Case B," involves integrating environmental, closure, and social costs within the block model. It is important to note that Deswik employs algorithms to solve scheduling and sequencing problems, considering temporal variables, operational constraints, and multiple destinations for extracted material. Additionally, the software incorporates algorithms for calculating the final envelope of the exploitation (Deswik, 2018; Poniewierski, 2018). Finally, using historical thermal coal prices, future price projections were derived. This will facilitate a sensitivity analysis, where three price scenarios corresponding to the 25th percentile (P25), 50th percentile (P50), and 75th percentile (P75) are proposed. These projections are generated using @Risk 8.2, based on the distribution function derived from the complete dataset. Fig. 1 summarizes and visually illustrates the proposed methodology for the development of this article. 2.1. Case study The case study involves a sedimentary coal deposit where an openpit mine is being developed in northern Colombia. At the request of the company and under a confidentiality agreement, the name of the company and the exact location of the project will not be disclosed; however, the provided data is part of an actual operation. The deposit is described as a syncline composed of 26 coal seams, which exhibit variable thicknesses along their extent, with pinch-outs in some areas. Like the thicknesses, the physicochemical characteristics are not continuous, which is why it is necessary to develop a block model with partial qualities, considering the average properties of the seams that are included in each block within the model. The dimensions used to create the block model are 10 m x 10 m x 10 m, which were grouped into macro-blocks of 50 m x 50 m x 50 m to reduce computational complexity. In Fig. 2, the morphology and topography of the sedimentary deposit can be observed. The deposit is projected to extend 4.5 km in length and 2 km in width at the surface. The gray strata represent the material of interest (coal), while the orange strata indicate waste material that will need to be disposed of in dumps, should it be extracted in order to reach the coal seams. Table 1 presents the main conditions and geological characteristics of the deposit in the case study. The initial costs that will be considered for both scenarios include D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 2 transportation costs and processing costs. Due to the variation in mining costs (which include transportation and blasting) based on block depth, an equation was developed as a function of Z, where the maximum value is 29.2 USD/t for the deepest blocks (Z = − 120) and the minimum value is 11.7 USD/t for the shallowest blocks (Z =230). The processing costs considered are related to coal washing to reduce the amount of ash; one of the assumptions made is that all the mineral will undergo a washing process. In addition to these costs, inputs such as waste and mineral mining capacities, the discount rate, mineral price, and final pit wall angle were also taken into account. Table 2 presents the considerations made for the case study. 2.2. Cost quantification The parameters suggested by Xu et al. (2014), for estimating environmental and ecological costs are summarized in Fig. 3, where the variables considered for each component can be observed: value lost from direct ecosystem services (Cd), value lost from indirect ecosystem services (Ci), and CO 2 emission costs (Cpr). In this article, the opportunity cost of developing agricultural activities will also be considered, as agriculture is one of the productive vocations in the study region. The closure plan cost is calculated based on reference rates (Units of Account, UF, in this case) and also takes into account data contributions from closure plan values in the mining industry. Since each closure case has its own specific characteristics, value corrections are made based on geographic location, altitude above sea level, and the distance to supply centers. The Unit of Account (UF) is a Chilean accounting unit created to adjust commercial and accounting transactions that change according to inflation. The value of the UF was assumed to be 37.44 USD. This methodology includes the costs mentioned in Law 20,551 and by Valdebenito (Valdebenito, 2015), which relate the economic values of the cubic measurements to the reference rate (UF) for each work or installation in the project. These data are presented in Table 3. The correction factors for this methodology are proposed by Valdebenito for the Chilean case; however, since the case study is in Colombia, it is suggested that: a) The geographical correction factor in the Chilean case varies between 0.91 and 1.0 depending on the region. *It is recommended that this value be modified based on the precipitation ranges suggested by the IDEAM (IDEAM, 2024), where areas with lower precipitation levels range from 0 mm to 500 mm annually, and areas with higher precipitation levels exceed 11,000 mm annually. Table 4 specifies the correction factor to be used for Colombia: b) The altitude correction factor is 1.0 for operations below 3000 m above sea level (m.a.s.l.), 1.25 for operations between 3000 and Fig. 1. Summary of the methodology proposed for the article. Fig. 2. Morphology and topography of the deposit. Table 1 Geological conditions and characteristics of the deposit. Parameter Value Unit Average ash 4,05 % Average coal density 1,31 t/m 3 Average calorific value 12,812 BTU/lb Average sulphides 0,75 % Number of macroblocks 25,060 – Annual extraction capacity 15 Mt Table 2 Initial conditions (inputs) for the sequencing of Case A. Parameter Value Unit Coal price 46 USD/t Mining cost -(0.05*Z) +23.2 UDS/t Wash cost 3.8 USD/t Annual waste removal capacity 115 Mt Discount rate 8 % Average slope angle 50◦– D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 3 4000 m.a.s.l., and 1.45 for operations above 4000 m.a.s.l. *For the present article, these values are modified based on the needs of the country. The altitude correction factor will be as follows: projects below 1000 m.a.s.l. will have a correction factor of 1.0, operations between 1000 m.a.s.l. and 2000 m.a.s.l. will have a correction factor of 1.25, and operations between 2000 m.a.s.l. and 3000 m.a.s.l. will have a correction factor of 1.45. In the country, extractive activities are not allowed in areas with altitudes greater than 3000 m.a.s.l. (classified as paramo zones). c) It is recommended to use the correction factor for supply centers (Fp) based on the distance from the mining operation to the nearest urban perimeter of the supply center (x) using Eq. (1): Fp=e0,0014 xif 0<x<650km (1) Fp=2,5if 650km ≤x Finally, the total closure costs can be calculated by correcting the direct costs found in the cubic measurements using the correction factors mentioned earlier, and adding the indirect costs and VAT, which in the case of Colombia is 19 %. To define the amount of money that should be invested as a result of conflicts with local communities, hypothetical values were assumed based on what the inhabitants would be willing to receive annually for their household as compensation for the mining project operation. Since several mining projects coexist in the study area, it was assumed that the number of affected inhabitants is proportional to the amount of mineral extracted by the project. All the costs calculated for this objective will be summed and unified into a single value, which will then be internalized in the mine planning process (calculation of the envelope, block sequencing, and scheduling). 2.3. Internalization of costs in the block model After determining Case A, the economic and operational values such as NPV, LOM, extraction rates, volumes, and the waste-to-mineral ratio (W/O) are calculated. To internalize the costs obtained through the previously mentioned methodologies, their equivalent value in terms of BTU/cal per metric ton was calculated. This allowed for the integration within the geological model, adjusting the calorific value of each of the blocks. A new scenario (Case B) was calculated where the mineral blocks are already associated with environmental, social, and mine closure costs in terms of their calorific value. The optimization and sequencing of this case will have different indicators that are compared with Case A. It is important to mention that, unlike other types of minerals, coal has several physicochemical properties such as sulfur, ash, and volatile percentages, among others, that it must meet for export purposes. To reduce the computational complexity of the problem, the study considers calorific value as the only characteristic parameter to optimize the mining plans, as the nature of thermal mineral coal makes this its main property. The assumed price of the metric ton of thermal coal is 11,370 BTU/lb for deliveries made in terms of ARA (Amsterdam, Rotterdam, Antwerp). Therefore, the thermal coal quotation price (Pton)must be adjusted by a correction factor related to its calorific value (Kc), which is based on a multiplying factor in terms of BTU/lb equivalents for the price of the metric ton (PBC). This is explained in Eq. (2): Pton =PBC ∗Kc 11,370 BTU/lb (2) The final pit calculation was performed using the pseudoflow tool in the Deswik mine planning software. Subsequently, with operational constraints and the characteristics of the operation, the sequencing and scheduling of the material to be extracted, which will be categorized as waste or mineral of interest, will be carried out. The objective of establishing the timing and destinations of the materials to be mined is to make the mine plan operational and identify areas that can be abandoned from the early stages of the project’s life in order to begin progressive closure activities. It is important to mention that the environmental costs of each block may vary depending on its specific chemical, physical, and spatial characteristics. For example, arsenic content can generate toxic drainage, pyrite or sulfide content can lead to acid drainage, and surface features such as different plant species or bodies of water can also influence costs. However, accurately redistributing the calculated costs non-uniformly among the blocks and determining the portion of these costs that should be allocated to mitigating a specific impact is challenging. For these reasons, and to reduce computational complexity in Fig. 3. Environmental costs taken into account in the article. Table 3 Values to be invested for each of the mine’s works and facilities. Parameter Unit Amount of UF per unit Steel buildings and structures ton 15 Concrete structures m 2 12 Industrial equipment ton 66 Camps and offices m 2 1,1 Pits, quarries and subsidence areas m 2 6 Waste dumps ha 186 Tailings deposits ha 379 Leachate ha 2628 Access roads m 0,3 Power lines m 0,3 Rail lines m 8 Coverages*ha 750–5000 Revegetation m 2 0,3 Soil remediation m 3 8 Hazardous industrial waste ton 8 * They are subject to climatic conditions; for areas with little rainfall such as deserts, approximately 750 UF/ha is recommended, and for areas with high rainfall, approximately 5000 UF/ha. D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 4 the block model sequencing, a uniform cost allocation was applied to each block. Fig. 4 graphically explains the proposed methodology for the internalization of costs into the geological model, allowing for sequencing with the included costs. 2.4. Sensitivity analysis Two of the methodologies that will be used to integrate price uncertainty are time series and probability distributions, which employ historical data to generate potential future scenarios. The recommended selection index to consider in such models is the Akaike Information Criterion (AIC), which is an asymptotic unbiased estimator between the proposed model and the true model (L´ opez, 2011). Using the "Time Series" tool in the software @Risk 8.2, it is possible to derive these time series through various simulations of random movements that take into account the behavior of thermal coal prices or any other mineral. Based on the results obtained in Case B, an analysis will be performed in which three hypothetical future price scenarios are proposed, corresponding to the 25th percentile (P 25), 50th percentile (P 50), and 75th percentile (P 75), using the price projection derived with @Risk 8.2. The coal prices used were taken from the year 1994 and are expressed in US dollars per metric ton. 3. Results 3.1. Case A 3.1.1. Technical aspects After completing the mining sequencing, results were obtained that describe the economic, production, and operational characteristics of Case A, which are shown in Figs. 5, 6, 7, 8, and in detail in Table A of the Appendix. It is important to mention that, up to this point, the Net Present Value (NPV) of Case A has not been affected by the mine closure costs, environmental costs, and costs related to social conflicts, which will be calculated in Sections 3.1.2, 3.1.3, and 3.1.4, respectively. In this case, the costs will be entered ex-post, so that once calculated, they will be subtracted from the NPV of this case to obtain the real NPV. As can be seen in Figs. 5 and 6, the rate of mineral extraction and waste removal remained constant throughout the life of the mine (LOM), which was 18 years. The only year that showed a substantial decrease in the extraction rate was year 3, which may be due to the morphology of the sedimentary deposit, where a larger amount of waste must be removed to access the coal seams. Additionally, the waste-to-mineral ratio remained above 19, which caused a drop in the NPV for that period of time. However, it is worth noting that this ratio also remained relatively constant, oscillating between 7.6 and 10.5 during the other 17 periods. In Fig. 7, this behavior can be analyzed graphically. The cumulative NPV is also shown in Fig. 8. Based on the amount of mineral and waste extracted, the final pit has a dimension of 1023 hectares, with a total waste production of 2.506 Table 4 Correction factor according to rainfall in the area or case study. Precipitation (mm per year) 0–500 5001000 1000–1500 1500–2000 2000–2500 2500–3000 3000–4000 4000–5000 5000–7000 7000–9000 9000–11,000 >11,000 Correction Factor 0.91 0.92 0.93 0.94 0.95 0.96 0.98 1.00 1.02 1.06 1.11 1.15 Fig. 4. Methodology for the inclusion of costs. D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 5 million tons (Mt), considering a swelling factor of 1.2. Of this total waste, 70 % is planned for disposal in waste dumps, while the remaining 30 % will be used for backfilling in abandoned areas. Therefore, for the construction of the waste dumps, an area of 1200 hectares is estimated, with a height of <70 m This configuration ensures that the waste disposal and backfilling processes align with the project’s environmental and operational plans, helping manage the waste effectively while also mitigating potential impacts on the surrounding environment. 3.1.2. Quantification of mine closure costs Following the proposed methodology, the stages for quantifying the closure costs were followed. Table 5 defines the cubic measurements for each of the variables related to mine closure costs. Each of these is associated with an estimated investment amount in Units of Account, the cubic measurements, and the corresponding investment amount in USD. Coverages(UF ha)=0.4∗(Average rainfall(mm year))+650 (3) It is important to clarify that the term "Cover" refers to the works that must be carried out to isolate tailings ponds, which are generally related to the waste from metallurgical processes. These are not applicable to the case study but were considered for the closure of sedimentation ponds following the coal washing plant. It was assumed that soil remediation would take place in areas where the waste rock is disposed of, and revegetation would occur in the areas corresponding to the final pit. Based on the values determined for the calculation of closure and abandonment costs, a total of 42,550,463 USD was obtained, to which the correction factors must be applied. In Table 6, the values considered for the application of each correction factor are described. Finally, the obtained value, which corresponds to the total amount that should be invested for closure and abandonment activities, is Fig. 5. Mineral extraction by period. Fig. 6. Extraction of waste (red) and mineral (blue) by period. Fig. 7. waste/mineral ratio by period. Fig. 8. Net Present Value accumulated throughout the life of the mine. Table 5 Assumed values of the variables for calculating mine closure costs. Parameter Total units Cost (USD) Steel buildings and structures 7.800 117.000 Concrete structures 60.000 720.000 Industrial equipment 20 1.320 Camps and offices 3.500 3.850 Pits, quarries and subsidence areas 1.023 6.138 Waste dumps 1.200 223.200 Tailings deposits 5 1.895 Leachate 0 0 Access roads 2.700 810 Power lines 5.000 1.500 Railways 0 0 Coverages (ha): They are subject to climatic conditions, for areas with little rainfall such as deserts approximately 750 UF/ha is recommended, areas with high rainfall around 5000 UF/ha. 5 5.750 Revegetation 10 ′ 230.000 3 ′ 069.000 Soil remediation 4 ′ 800.000 38 ′ 400.000 Hazardous industrial waste 0 0 TOTAL 42′550.463 Note: To calculate the amount of UF for the vegetation cover parameter, the average value of the lower range (250) and the upper value (11,000) from the scale shown in Table 4 was used. Based on these values and the recommendations in the Law, which suggest investing 750 UF/ha for areas with low precipitation and 5000 UF/ha for other areas, Eq. (3) is proposed to determine the value of the Cover parameter (ha). D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 6 56,905,665 USD. 3.1.3. Quantification of environmental costs According to data reported by the Institute of Hydrology, Meteorology, and Environmental Studies – IDEAM (IDEAM, 2024), the conditions under which the mining project in the case study is developed correspond to a terrestrial ecosystem intended for agricultural activities, specifically a mosaic of pastures and natural spaces typical of the northern region of Colombia. There are no significant watercourses in the mining area, only seasonal streams with low flow during the rainy seasons (October and November), and annual precipitation varies between 1000 and 1500 mm. In Table 7 the environmental variables and the values used for each of them are presented,v* was calculated based on the net annual income per hectare for maize crops in Colombia, with an average price of 256 USD/t. Similarly, for calculating Cd, the opportunity cost lost by maize crops in the final pit area was considered, with a yield of 38.1 tons of maize per hectare and three harvests per year. It is important to mention that the average gasoline consumption (eg) was not considered, as all the equipment used in the case study operates on diesel fuel, and the electricity used in the project does not come from thermal coal. Therefore, b, ac and ee are assigned a value of 0. The total environmental costs. As shown in Table 7, the total environmental costs that the project will incur amount to 22,179,596 USD. 3.1.4. Quantification of costs of resolving social conflicts The number of inhabitants in the populations directly affected by the mining operations adjacent to the urban centers or small villages in the study area is 60,000, who are directly or indirectly influenced by the coal extraction projects taking place there. To calculate the number of inhabitants affected by the project in the case study, a proportional relationship was established between the amount of mineral extracted in the last year in the mining area and the amount extracted by the operation under analysis. Thus, the projected annual production is 15,000,000 tons of coal, while the average production of all the projects in the region was 80,215,000 tons of coal, accounting for 18.7 % of the total production in the mining area (11,220 people). Additionally, it was assumed that the average number of people per family unit is 6, and the average amount of money each family is willing to receive is 827 USD/ year. Therefore, a total of 1870 family units will receive the compensation, equivalent to 1546,478 USD/year and 27,836,602 USD for the entire life of the mine. This latter value will represent the total social costs for the case study. The total calculated costs amount to 106,921,863 USD, which corresponds to the total costs that had to be calculated for the case study. If this value is subtracted from the 1060,780,000 USD corresponding to the net present value (NPV) of the sequencing of Case A, the result is 953,858,137 USD, which represents the real NPV of this case. It is necessary to clarify that these values are deducted directly from the NPV in year 0 and are not affected by the discount rate, as they are generally paid before the start of the project as a financial guarantee, depending on the legal framework in which the mine operates. 3.2. Internalization of costs in mining optimization (Case B) Once Case A has been identified with its economic and operational characteristics, such as NPV, LOM, extraction rates, volumes, and the waste-to-mineral ratio already determined, it is possible to internalize the calculated costs for the mining sequencing by calculating their equivalent value in BTU/cal of metric tons. According to the methodology proposed, the costs to be internalized in the mining blocks will include the total costs of closure and abandonment, environmental costs, and conflict resolution costs, which amount to 106,921,863 USD. It is therefore necessary to calculate their equivalent in BTU/lb to account for the quality of the coal to be extracted. The value used is the one mentioned in Eq. (3), where the assumed cost of 46 USD/ton has a calorific value of 11,370 BTU/lb, resulting in a total of 1215,701,582,310 BTU/lb to be internalized and incorporated into the model. The calculated BTU/lb will be distributed to the blocks extracted in Case A by decreasing the coal quality. This will allow the sequencing algorithm of the software to reduce the economic value of each block and make new decisions regarding its destination (plant or waste dump) and the period in which it will be extracted. It is important to mention that reducing the coal quality will change the economic and technical conditions of the project, such as the amount of waste and mineral to be extracted, the waste-to-mineral ratio, the NPV, and the LOM. However, these results would already include the Table 6 Final calculation of closure and abandonment costs taking into account correction factors. *The assumed value for the correction factor a, corresponding to the correction factor for the urban center, is 5 km. Initial calculation 42′550.463 USD Correction Factor Value Value a 0,93 39 ′ 571.930 USD b 1 39 ′ 571.930 USD c 1007 39 ′ 849.905 USD Others Indirect costs 0,20 47 ′ 819.887 USD VAT 0,19 56 ′ 905.665 USD Table 7 Assumed values for calculating environmental costs. Symbol Meaning Unit Value VsSoil erosion control $ 71,71 sSoil retention capacity t/hm 2 /a 42,50 v* Annual net soil income from local crops $/hm 2 9.753,10 ρ sSoil density for cultivation t/m 3 1,30 hmSoil thickness m 0,40 VaAir pollutant uptake $ 1694,82 qNet primary productivity ton/(Ha*yr) 38,10 fcCarbon mass conversion factor C to CO 2 dimensionless 3,67 cCCarbon capture and storage cost $/t 65 ysSO 2 absorption capacity of vegetation t/hm 2 /a 0,15 csSO 2 removal and control $/t 114 ydDust absorption capacity of forest t/hm 2 /a 21,65 cdDust removal and control cost $/hm 2 27 VoOxygen release $ 182,88 coOxygen extraction cost $/t 4 VrRunoff control $ 9900 pAverage annual rainfall (mm/a) 1250 kProportion of rainfall causing runoff dimensionless 0,30 frVegetation-caused runoff flow reduction coefficient dimensionless 0,24 crCost of storing water in a reservoir $/m 3 11.000 VnSoil nutrient formation $ 67,38 kNSoil nitrogen content. dimensionless 0,0043 PNNitrogen fertilizer cost $/t 211 fpP to P 2 O 5 mass conversion factor dimensionless 2,2903 kpSoil phosphorus content dimensionless 0,00,037 PpPhosphorus fertilizer cost $/t 113,82 kKSoil potassium content dimensionless 0,00,242 PKPotassium fertilizer cost $/t 316 CiIndirect eco-services lost value (Sum of the above V) $ 11.924,79 CdDirect eco-services lost value (Opportunity costs) $ 9.753,1 edAverage diesel consumption per tonne of material removed kg/t 0,67 fdCarbon emission factor for diesel dimensionless 0,87 cCCarbon capture and storage cost $/t 75 Cpr CO 2 emission costs for energy consumption $ 0.4 TOTAL ($/ha) 21.680,93 TOTAL ($) 22 ′ 179.595,78 D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 7 costs calculated in the previous section, thus significantly reducing the risks associated with incorporating them ex-post. Figs. 9, 10, 11, 12, and the detailed Table B in the Appendix show the results obtained after internalizing the costs and performing the block sequencing for Case B. 4. Discussion The results obtained in the cost internalization case show considerable variations in the waste-to-mineral ratio, the amount of waste extracted, NPV, and LOM. It is important to note that in both cases, the number of mineral extracted was similar—253.47 Mt for Case A and 251.74 Mt for Case B. This is significant as it confirms the amount of reserves present in the deposit and indicates that the inclusion of costs did not drastically affect this value in our case study. Although similar extraction levels were obtained, the extraction rate experienced notable changes, which are important to analyze. First, it can be observed that the sequencing process provided the software with more flexibility in the extraction rate, suggesting to the user that extraction periods should exceed 15 Mt for years 9 and 10. If this "flexibility" were not implemented during these two periods, the NPV would likely be severely impacted. This could be explained by the morphology of the deposit, as it is necessary to extract a bit more in these years to reach the next coal seams. Otherwise, extracting them in subsequent periods may not be economically viable, also affecting future production. It is important to emphasize that this increase in the extraction rate represents only 3 % of the 15 Mt initially considered. Secondly, annual production in Case B showed fewer variations than in Case A, which had reduced extraction for years 3 and 9, with values of 6.01 Mt and 10.94 Mt, respectively. Case B only presented significant variations in year 16, with an extraction of 13.44 Mt of coal. Another factor that can be analyzed is the amount of waste extracted. This is crucial as it can represent changes in the project’s environmental performance. As noted in earlier sections, if more waste needs to be removed, the associated rehabilitation and closure costs will also increase. In both cases, the extraction rate is constant and fluctuates within the same values. However, in Case A, 2088.61 Mt of waste was extracted, while in Case B, 1970.73 Mt was extracted. This could be explained by the reduction in LOM, which was shortened by one year, meaning there would be no need to remove the mineral from year 18, as continuing extraction would not be economically viable. The reduction in LOM and the amount of extracted mineral indicates that extracting more mineral, even if it is in situ, is not always the most cost-effective strategy. In fact, extracting greater quantities of mineral could be counterproductive for the project’s NPV, as it will tend to decrease. Another factor indicating better environmental performance is the waste-to-mineral ratio, as it shows that less waste is required to obtain the same amount of coal. This was one of the indicators that improved in comparison to Case A, as the average value decreased from 8.24 to 7.84 (approximately 5 %) and also showed more stable behavior (with no value exceeding 8.56), unlike in Case A, where periods with a waste-tomineral ratio of 19.5 and 10.51 were observed. Lastly, and as expected, the NPV of Case B showed a reduction compared to Case A. Despite the NPV of 898.30 million USD being 4 % lower than the initially calculated value, this scenario presents lower environmental and social risks and uncertainties than the initial case. Therefore, the potential environmental issues and community conflicts would already be integrated into the mining plans. This is of utmost importance as it guarantees that the chances of suspending or prematurely closing the project are reduced. It is important to remember that this is one of the main challenges faced by mining projects around the world, especially in Latin America, where various social movements have influenced the suspension or closure of projects that were either already operating or about to start. Fig. 9. Mineral extraction by period in Case B. Fig. 10. Extraction of waste and mineral by period in Case B. Fig. 11. Waste/Mineral ratio by period in Case B. Fig. 12. Accumulated Net Present Value, with internalized costs, throughout the life of the mine. D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 8 4.1. Sensitivity analysis Due to the uncertainty associated with the variables involved in the results of optimization processes in mining plans, the future variation of some of these variables, especially the prices of the minerals to be extracted, must be considered. Two of the methodologies used to integrate price uncertainty are time series and probability distributions, which employ historical data to generate possible future scenarios (Franco Sepúlveda et al., 2012). The recommended selection criterion for this type of model is generally the Akaike Information Criterion (AIC), which is an asymptotic unbiased estimator between the proposed model and the true model (L´ opez, 2011). Using the "Time Series" tool of @Risk 8.2 software (PALISADE, 2023), it is possible to generate these time series through different simulations of random movements that account for the behavior of the copper price in this case. Considering the results obtained in Case B, three hypothetical future price scenarios are proposed corresponding to the 25th (P25), 50th (P50), and 75th (P75) percentiles, obtained through a price projection with @Risk 8.2. The coal prices used were taken from a time period starting in 1994 and are expressed in USD per ton. The function used by the @Risk 8.2 software was RiskARCH1 ( μ , ω , b 1 , Y 0 ), which "generates a first-order autoregressive conditional heteroscedasticity process" (PALISADE, 2023) and is recommended for predicting the future price of commodities. The operation of this function is expressed in Eqs. (4) and 5, which are presented below: Eq. (4). From Palisade, 2023 Yt= μ + σ tNt(4) When σ t is modeled as follows: σ 2 t= ω +a1(Yt−1− μ ) And: Nt: is a normal distribution function (0,1) μ =37,239 (average) ω =244,3 (variance equation constant) b 1 =0,57,735 (coefficient of error), Y 0 =46 (time value 0) In Figs. 13 and 14 it is possible to see the behavior of the projections made, and the distribution function in which the 25th (P 25) and 75th (P 75) percentiles are identified, with their respective values. In the graphs shown, it is possible to identify that for the 25th (P25) and 75th (P75) percentiles, the values are 26.6 USD/t and 47.9 USD/t, respectively. In the case of the 50th percentile (P50), this corresponds to the mean of the function, which was 37.24 USD/t. These values were considered for performing the sensitivity analysis, where the coal price will essentially replace the value in the economic valuation of the mine sequencing. It is important to emphasize that, to reduce the amount of data, only the relevant results from the identified scenarios will be shown (total mineral extracted, total waste extracted, accumulated NPV, LOM, and waste-to-mineral ratio). Table 8 presents the results of the sensitivity analysis with the three proposed scenarios. The sensitivity analysis conducted on the coal price indicates that the project is vulnerable to drastic changes if low mineral prices are considered. The scenarios for the 25th and 50th percentiles suggest that there is a possibility that the project may not be viable under the conditions proposed, not so much due to the NPV, as both cases are above 0, but rather due to the short mining life. With such low prices, the software algorithm only finds mineral extraction to be profitable for periods of 2 and 5 years, respectively. Such short LOMs are not feasible due to the large investments required to start a project, which were not accounted for in this article (e.g., equipment acquisition, facility construction, personnel hiring, geological studies, etc.). The 75th percentile scenario is not shown in Table 8, as the software could not sequence the mined blocks with a higher cost. This was not due to profitability concerns, but rather because, under the given conditions, extraction capacities or periods need to be increased to find the optimal solution, which could undoubtedly lead to a higher NPV. However, increasing extraction rates, LOM, or changing other restrictions would make it impossible to compare with the case study, as it would represent a completely different project with other associated environmental costs, which would be incorrect to compare. Considering that projects with lower extraction rates and shorter LOMs require smaller initial investments, we can conclude that with lower prices, it is possible to propose smaller-scale productive alternatives that may still be profitable. Similarly, as shown in the 75th percentile scenario, it would be necessary to increase the restriction on extraction rates in order to achieve a project that aligns with the material handling capacity requirements. Fig. 13. Price projections using @Risk 8.2 software. Fig. 14. Price projection distribution function. Table 8 Results of price sensitivity analysis. Case B Scenery P25 Scenery P50 Scenery P75 Mineral Extracted (Mt) 251,57 22,65 63,89 * Waste Extracted (Mt) 1.970,74 135 580,61 * NPV Accumulated (M USD) 898,30 178,43 331,20 * LOM (years) 17 2 5 * Waste/mineral ratio 7,84 7,87 9,1 * D. Oliveros-Sepúlveda et al. The Extractive Industries and Society 23 (2025) 101668 9