Advances in Mine Tailings and Water Management in the Mining Industry for a Circular Economy

ADVANCES IN MINE TAILINGS AND WATER MANAGEMENT IN THE MINING INDUSTRY FOR A CIRCULAR ECONOMY

Natalia Andrea Araya Gómez

Thesis presented in accordance to the requirements to obtain the degree of Ph.D. in Mineral Process Engineering

Supervisors Luis Cisternas Ph.D. - Andrzej Kraslawski Ph.D.

Laboratory of Optimization and Modelling Department of Chemical and Mineral Process Engineering Universidad de Antofagasta Antofagasta, Chile

2020

AVANCES EN EL MANEJO DE RELAVES Y DE AGUA EN LA INDUSTRIA MINERA PARA UNA ECONOMÍA CIRCULAR

Natalia Andrea Araya Gómez

Tesis para optar al grado de Doctor en Ciencias de la Ingeniería de Procesos de Minerales

Profesor Patrocinante Luis Cisternas Ph.D.

Profesor copatrocinante Andrzej Kraslawski Ph.D.

Laboratorio de Optimización y Modelamiento Departamento de Ingeniería Química y Procesos de Minerales Universidad de Antofagasta Antofagasta, Chile

2020

Supervisors

Professor Luis Cisternas Arapio Departamento de Ingeniería Química y Procesamiento de Minerales Universidad de Antofagasta Chile

Professor Andrzej Kraslawski Industrial Engineering and Management LUT School of Engineering Science Lappeenranta-Lahti University of Technology Finland

Abstract

Natalia Araya Gómez

The mining industry consumes vast quantities of water and energy to produce a metal or mineral product from mineral resources on the earth’s crust, leaving an enormous amount of mining waste.

The linear thinking of the economy needs to be replaced with the circular economy to achieve sustainable development goals. In this context, the mining industry needs to improve in several aspects, such as reducing primary resource consumption and improve the efficiency of its processes. Strategies to reduce water and energy consumption, recycle water and wastes, and to recover energy are needed. The objective of this thesis is to develop methodologies to advance towards the circular economy in mining, with a focus on mine tailings and water management. The strategies and tools proposed are meant to be applied in mining processes to mitigate environmental impacts. The strategies proposed in water management include an integrated water distribution network to supply water to several mine plants while recovering energy using energy recovery devices. The method used is mathematical optimization to design the water distribution network. The methodology is validated with a case study that corresponds to an area of the Antofagasta Region. Results show that the optimal solution is an integrated system with energy recovery devices in areas with a complex topography where energy production is feasible.

Reducing waste is a key component of the circular economy. Mine tailings are the main waste produced by the mining industries. Re-processing of mine tailings to obtain critical materials is proposed, the methodology is an economic assessment validated with a case study of tailing deposits of the Antofagasta Region. The economic evaluation considers the discounted cash flow method, sensitivity analysis, and real options analysis. Results show that an investment based on re-processing mine tailings to obtain critical materials is feasible in some cases.

The contribution of this thesis is a collection of methodologies to improve mine tailings and water management to improve mining processes. About mine tailings management, the strategy proposed is re-processing mine tailings, which will reduce the amount of waste produced by mining plants and will reduce the need for mining primary ores. In the case of water management, the mining industry needs to reduce the demand for freshwater. To supply seawater is still an expensive option but having an integrated water distribution network with energy recovery devices will reduce the total cost of the network.

Keywords: Water Distribution Networks, water management, mine tailings, mining, circular economy, critical materials, resource valorization, economic assessment

Resumen La industria minera consume grandes cantidades de agua y energía para producir un producto metálico o mineral a partir de recursos minerales en la corteza terrestre, dejando una enorme cantidad de desechos mineros.

El pensamiento lineal de la economía necesita ser reemplazado por la economía circular para lograr los objetivos de desarrollo sostenible. En este contexto, la industria minera necesita mejorar en varios aspectos, como reducir el consumo de recursos primarios y mejorar la eficiencia de sus procesos. Se necesitan estrategias para reducir el consumo de agua y energía, reciclar agua y desechos, y recuperar energía.

El objetivo de esta tesis es desarrollar metodologías para avanzar hacia la economía circular en la minería, con un enfoque en los relaves mineros y la gestión del agua. Las estrategias y herramientas propuestas están destinadas a aplicarse en los procesos mineros para mitigar los impactos ambientales de estos.

Las estrategias propuestas en la gestión del agua incluyen una red integrada de distribución de agua para proveer de agua a varias plantas mineras mientras se recupera energía utilizando dispositivos de recuperación de energía. El método utilizado es la optimización matemática para diseñar la red de distribución de agua. La metodología fue validada con un caso de estudio que corresponde a un área de la Región de Antofagasta. Los resultados muestran que la solución óptima es un sistema integrado con dispositivos de recuperación de energía en áreas con una topografía compleja donde la producción de energía es factible. Reducir la cantidad de residuos es un componente clave de la economía circular. Los relaves mineros son los principales desechos producidos por las industrias mineras. Se propone el reprocesamiento de relaves mineros para obtener materiales críticos, la metodología consiste en una evaluación económica validada con un caso de estudio de relaves mineros de la Región de Antofagasta. La evaluación económica considera el método de descuentos de flujos, el análisis de sensibilidad y el análisis de opciones reales. Los resultados muestran que en algunos casos es factible una inversión basada en el reprocesamiento de relaves de minas para obtener materiales críticos.

La contribución de esta tesis es una colección de metodologías para mejorar la gestión de los relaves mineros y la gestión del agua en los procesos mineros. Acerca de la gestión de relaves mineros, la estrategia propuesta es reprocesar los relaves mineros, lo que reducirá la cantidad de relaves producidos por las plantas mineras y reducirá la necesidad de extraer minerales primarios. En el caso de la gestión del agua, la industria minera necesita reducir la demanda de agua dulce. Suministrar agua de mar sigue siendo una opción costosa, pero tener una red integrada de distribución de agua con dispositivos de recuperación de energía reducirá el costo total de la red.

Acknowledgment I would like to express my deepest appreciation to Professor Luis Cisternas for being my supervisor and for encouraging me to pursue a double degree with LUT University. I must also thank the project Anillo Agua de Mar Atacama ACT 1201 - ANID in which I started my thesis research. I’m also grateful of the collaboration developed in the project Anillo Grant ACM 170005 - ANID.

I would also like to extend my deepest gratitude to Professor Teófilo Graber as head of the Ph.D. program in Mineral Process Engineering and to all members of Ph.D. in the Mineral Process Engineering program for all the support during the duration of my Ph.D. studies.

I would also like to express my deepest gratitude to Professor Andrzej Kraslawski, my supervisor at LUT University, for support during the realization of the double degree for my Ph.D. studies.

I would also like to extend my gratitude to Professor Edelmira Gálvez for her support and collaboration in my firsts two years of my Ph.D. studies in project Anillo Agua de Mar Atacama ACT 1201-ANID.

I am also grateful to Freddy Lucay for support and collaboration in my research while being in the Project Anillo Agua de Mar Atacama, resulting in two publications.

I am extremely grateful to ANID for the grant (2017, No. 21170815) for pursuing doctoral studies.

I am also grateful to Ella and Georg Ehrnrooth Foundation for a grant to continue my doctoral studies in LUT University.

List of publications

Publication I Araya, N., Lucay, F. A, Cisternas, L. A., Gálvez, E. D. (2018). Design of desalinated Water Distribution Networks: Complex Topography, Energy Production, and Parallel Pipelines. Industrial and Engineering Chemistry Research, 57, 9879-9888. DOI: 10.1021/acs.iecr.7b05247. Publication II

Araya, N., Kralawski, A., Cisternas, L. A. (2019). Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings. Journal of Cleaner Production, 263, 121555. DOI: 10.1016/j.jclepro.2020.121555 Publication III (submitted)

Araya, N., Rámirez, Y., Kraslawski, A., Cisternas, L. A. (2019) Real options valuation of processing mine tailings to obtain critical materials. Sustainable Development (submitted).

Book Chapters

Araya, N., Cisternas, L. A., Lucay, F. A., Gálvez, E. D. (2017). Design of desalinated water distribution networks including energy recovery devices. In Computer Aided Chemical Engineering Vol. 40, pp. 925-930. Elsevier.

Herrera, S., Araya, N., Cisternas, L., Gálvez, E. (2016) Diseño de plantas desalinizadoras y redes de distribución de agua: una mirada holística. “Agua de Mar Atacama: Oportunidades y avances para el uso sostenible del agua de mar en minería”. Capítulo 3. RIL Editores. ISBN: 978-956-01-0388-8.

Preface

This thesis titled Advances in mine tailings and water management for a circular economy presents the work performed during the Ph.D. studies started in April 2015 and finished in June 2020. This Ph.D. thesis was developed under a joint supervision agreement between Universidad de Antofagasta in Chile and LUT University in Finland. This document is arranged in two sections. The first section contains the thesis overview divided into six chapters. The first chapter is the introduction and provides a context about the background of the study and the motivation and scope of this thesis. The second chapter, theoretical background, gives an overview of the literature context. The third chapter contains the research methodology and methods developed during the study. The fourth chapter contains a review of the results and publications obtained during the duration of the Ph.D. studies. The fifth chapter is Conclusion, and it summarizes the theoretical contribution of this Ph.D. thesis, it also provides the limitation found and future suggestions. Chapter six contains the references. The second section includes the articles published during the Ph.D. study. The first article, “Design of desalinated water distribution networks: Complex topography, energy production, and parallel pipelines” was published in the Journal Industrial and Engineering Chemistry Research. The second article titled “Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings” was published in the Journal of Cleaner Production. The third article, currently submitted in Sustainable Development, is titled “Feasibility of re-processing mine tailings to obtain critical materials using real options analysis.”

Prefacio

La tesis titulada Advances in mine tailings and water management for a circular economy presenta el trabajo realizado durante los estudios de doctorado que comenzaron en abril de 2015 y terminaron en junio 2020. Los estudios de doctorado fueron realizados bajo un convenio de doble graduación entre la Universidad de Antofagasta en Chile y LUT University en Finlandia. Este documento está dividido en dos secciones. La primera sección contiene una visión general dividida en seis capítulos. El primer capítulo es la introducción, la que provee un contexto sobre los antecedentes del estudio, la motivación y objetivos de la tesis. El segundo capítulo, antecedentes generales, proporciona una visión sobre el contexto de la literatura. El tercer capítulo contiene la metodología de investigación y los métodos desarrollados durante los estudios de doctorado. El cuarto capítulo contiene la revisión de los resultados y las publicaciones realizadas durante los estudios de doctorado. El quinto capítulo, las conclusiones, resume las contribuciones teóricas de la tesis de doctorado, también proporciona las limitaciones y sugerencias para investigaciones futuras. El sexto capítulo contiene las referencias. La segunda sección contiene los artículos publicados durante los estudios de doctorado. El primer artículo “Design of desalinated water distribution networks: Complex topography, energy production, and parallel pipelines” fue publicado en la revista Journal Industrial and Engineering Chemistry Research. El Segundo artículo titulado “Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings” fue publicado en la revista Journal of Cleaner Production. El tercer artículo, fue enviado a la revista Sustainable Development tiene como título “Feasibility of re- processing mine tailings to obtain critical materials using real options analysis”.

Contents

Abstract...... 1 Resumen ...... 2 Acknowledgment ...... 3 List of publications ...... 4 Book Chapters ...... 4 Preface ...... 5 Prefacio ...... 6 SECTION I: DISSERTATION OVERVIEW ...... 9 1. Introduction ...... 10 1.1. Background ...... 10 1.2. Motivation and Scope ...... 11 2. Theoretical Background ...... 14 2.1. Circular economy and sustainability in mining ...... 14 2.2. Water supply for mining ...... 15 2.3. Mine tailings ...... 16 3. Research Methodology ...... 18 3.1. Water Distribution Network (WDN) design for areas with complex topography considering the energy recovery devices ...... 18 3.1.1. Optimization of Water Distribution Networks (WDN) ...... 18 3.1.2. Problem Statement ...... 18 3.1.3. Mathematical Formulation ...... 19 3.2. An economic evaluation of re-processing of mine tailings to obtain critical materials ...... 19 3.2.1. Project valuation Tools ...... 19 3.2.2. Discounted Cash Flow (DCF) ...... 20 3.2.3. Real Options Analysis (ROA) ...... 20 3.2.4. Sensitivity Analysis...... 21 4. Publications and Results Reviews ...... 22 4.1. Publication I: Design of desalinated water distribution networks: Complex topography, energy production, and parallel pipelines ...... 22 4.1.1. Research Objective ...... 22 4.1.2. Contributions ...... 22

4.2. Publication II: Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings ...... 23 4.2.1. Research Objective ...... 23 4.2.2. Contributions ...... 23 4.3. Publication III: Feasibility of re-processing mine tailings to obtain critical materials using real options analysis (submitted in Science of The Total Environment) ...... 23 4.3.1. Research objective ...... 23 4.3.2. Contributions ...... 24 5. Conclusions ...... 25 5.1. Theoretical Contribution of the Study ...... 27 5.2. Limitation and Future Research Suggestions ...... 27 References ...... 29 SECTION II: PUBLICATIONS ...... 35

SECTION I: DISSERTATION OVERVIEW

1. Introduction 1.1. Background Mining is water and energy-intensive activity, an industry that is needed to ensure the wellbeing of modern society, moreover mining is present in civilization since ancient times. Mining activities are usually located in remote and arid areas where water resources are scarce (Cisternas and Gálvez, 2018; Herrera-León et al., 2019; Ramírez et al., 2019). Environmental and social awareness in the last decades around the use of resources has increased. Nowadays, terms like “circular economy” and “sustainability” resonate in different fields, mining is no exception (Adibi et al., 2015; Caron et al., 2016; Dong et al., 2019; Gorman and Dzombak, 2018; Lottermoser, 2011; Reyes-Bozo et al., 2014). A circular economy is an economy that is restorative and regenerative by design, using principles of reusing, recovery, recycling, it is about looking beyond the take-make- waste extractive industrial model (Kirchherr et al., 2017). The concept of circular economy recognizes the need to work on all scales, while sustainability focuses on three pillars: economic, environmental, and social, to meet the demands of the present, ensuring the ability of future generations to also meet their needs.

In Chile, there are numerous mineral processing plants, since Chile is the world´s first producer of copper. Other metallic elements extracted in Chile are silver, molybdenum, and gold. As well as for non-metallic elements as boron, iodine, lithium carbonate, among others (SERNAGEOMIN, 2019). Most of the mining facilities in Chile are located in the Atacama Desert in Northern Chile, as it features plenty of minerals. Mining is the principal activity in this area. Usually, plants are located away from the seashore, so in case of using desalinated water, piping water through long distances is a must (Herrera-León et al., 2019). Seawater use in mining has been growing since mines need to reduce the amount of raw water intake from groundwater and freshwater sources since these are limited and scarce (Cisternas and Gálvez, 2018). The methodologies developed in this thesis are applied to study cases based on the mining industry in Northern Chile, specifically the copper mining activities develop in the Antofagasta Region.

Reducing water consumption and increasing energy efficiency are two essential requirements to have a more sustainable mining industry. A typical base-metal mine site consists of an underground or open-pit mine and a mineral processing plant or mill (Gunson et al., 2010). Most mills use wet processes such as froth flotation to separate valuable minerals from the non-valuable minerals in the ore. Wet processes require less energy, but they can use large quantities of water (Gunson et al., 2010).

Currently, each plant has built its desalination plant and water distribution system. Since mining plants are located far away from the seashore, a way of reducing the energy used for pumping desalinated water through different mine sites is to have an integrated water supply system, where industrial symbiosis can be achieved between desalination companies and mining plants (Cisternas and Gálvez, 2018; Herrera-León 10 et al., 2019; Ihle and Kracht, 2018). Energy can be saved by using an integrated water supply system, efficiency can be achieved by using fewer pipelines and pump stations if energy recovery systems are placed in the pipelines, energy can be produced and used in the network (Corcoran et al., 2016; McNabola et al., 2014; Nogueira-Vilanova and Perrella-Balestieri, 2014a, 2014b; Zakkour et al., 2002).

Another issue in mining is mine tailings, which are the waste produced in mining processing plants. They are a mixture of fined solid materials remaining after the recoverable metals and minerals have been extracted from mined ore, together with water, which includes also dissolved metals and ore processing reagents (Edraki et al., 2014). Mine tailings can account for 95-99% of the crushed and ground ores (Edraki et al., 2014).

Environmental impacts and land use aspects should be considered when choosing a tailing disposal method (Adiansyah et al., 2017). Environmental impacts of tailings dam failures include human fatalities, contamination of water, and soil (Kossoff et al., 2014). One of the most severe environmental problems associated with mine tailings is acid mine drainage, which is produced when sulfide minerals are exposed to oxygen and water (Akcil and Koldas, 2006). Acid mine drainage is one of the most significant forms of water pollution (Moodley et al., 2017). Mine tailings management is crucial in mining operations because of the irreversible impacts of tailings (Adiansyah et al., 2015).

Mine tailings contain several elements that are not separated from the mineral in the processing plant. Mine tailings could be reprocessed to obtain valuable elements contained in them (Binnemans et al., 2015; Falagán et al., 2017). The valorization of mine tailings is still at its infancy, globally speaking (Kinnunen and Kaksonen, 2019). Critical materials are a group of materials that are considered of large economic importance and have high risk in their supply. Now, secondary sources like metal scrap and mine tailings have a promising future as critical materials sources. Using secondary sources to recover critical materials has multiple benefits: it allows to reduce primary ore extraction, decrease the amount of waste, hence decreasing its environmental impact, give value to waste, contributes to industrial symbiosis, among others. 1.2. Motivation and Scope A requirement to achieve sustainable development is to reduce primary resource consumption, consequently mining activities on primary sources. However, the demand for metals will continue to exist throughout the transition to a more sustainable society. Therefore, research must be conducted to explore the role of the mining industry in the circular economy paradigm and identify strategies for it to transitioning to a more sustainable mining industry.

In mining, circular economy advocates for reducing primary resource extraction by replacing it with secondary sources, reducing raw water intake by reusing or recycling

11 water, reducing energy dependency, reducing the environmental footprint, reducing greenhouse gas emissions (Lèbre et al., 2017).

Effective water management in the mining industry can reduce the dependence on freshwater sources that are scarce in arid regions where most of the mining facilities are located. In 2018, in Chile freshwater consumption reached 13.36 m3/s, more than three times seawater consumption, which reached 3.99 m3/s (COCHILCO, 2019a). Integrated water supply systems between desalination plants and several industries reduce energy consumption (Herrera-León et al., 2019) and the stress on freshwater resources. Different strategies can be applied to reduce water consumption and optimize its use. Mine tailings are a critical issue in mining as tailings are the waste of processing plants and are produced in large quantities. In Chile, to March of 2018, 740 tailing deposits are registered in a national cadaster, being the majority located in the north of the country. Currently, 101 tailing deposits are in active use, which will accumulate 14,470 million cubic meters, of the 16,000 million cubic meters of the total of tailings stored in the country, ergo 90.6% of stored tailings come from active mining operations (SERNAGEOMIN, 2018).

Strategies to save water, recover energy, and reducing waste are needed to make the mining industry more sustainable and to help to achieve the circular economy. The motivation of this thesis is to elaborate research to contribute to the literature on water and tailings management in the mining industry by providing methodologies and tools that mitigate the environmental impact of mining activities.

This thesis aims to propose strategies to manage water and tailings in the mining industry. Optimized water distribution networks and efficient water use in tailing management are proposed as water management strategies. Recovery of critical materials and reducing water content in tailings are the tailings management strategies proposed as well. The aim is attained by fulfilling the following specific objectives and the related research questions: 1. To design integrated water supply systems that include energy recovery devices in areas with complex topography RQ 1.1. Is an integrated water supply system that includes energy recovery devices a feasible option in areas with complex topography? RQ 1.2. Is it economically and technically feasible to use energy recovery devices in the water supply system? RQ 1.3. When using parallel pipes is a feasible option in the water supply system?

2. To investigate the feasibility of re-processing mine tailings to obtain critical materials RQ 2.1. What critical materials can be recovered from mine tailings?

RQ 2.2. What are the challenges in the production of critical materials using mine tailings as a source? RQ 2.3. Is it economically feasible to invest in a project developed around the idea of re- processing mine tailings to obtain critical materials?

3. To develop a methodology to assess the feasibility of re-processing mine tailings to obtain critical materials RQ 3.1. How can real options improve the decision to invest in a project using mine tailings as a source of critical materials? RQ 3.2. How can the net present values variables influence real options performance?

2. Theoretical Background Building on the literature review, this section explains the main concepts used in the thesis and the current situation of the mining industry encompassing the concepts and definitions used in this thesis. The demand for copper expected to increase and to exceed copper mineral resources by mid-century. The increase in copper demand will be associated with a strong increase in the energy demand and, consequently, the environmental impacts (Elshkaki et al., 2016). Environmental impacts associated with mining are stress and pollution of water resources, air pollution, land usage, and production of several waste streams.

In Chile, mining is a long-standing industry, since the country owns vast mineral resources. Most of the mining activity is located in the Atacama Desert, northern Chile, where sulfides ores containing copper are abundant; additionally, non-metallic resources are also found in the area. Therefore, Chile produces various elements, such as copper, silver, molybdenum, lithium, boron, among others (SERNAGEOMIN, 2019). Mining represents 10% of the national Gross Domestic Product (Reporte Minero, 2018). To achieve sustainable mining, the Chilean state must balance economic growth with the strictness of environmental policies by continuing to address the challenge of energy use, water scarcity, indigenous rights, and governmental organization (Ghorbani and Kuan, 2017). 2.1. Circular economy and sustainability in mining The circular economy is a concept that resonates lately, which describes an economic system that is based on business models that replaces the linear economy thinking, with reducing, alternatively reusing, recycling, and recovering materials in production/distributions and consumption processes, thus operating at all levels, to accomplish sustainable development, which implies creating environmental quality, economic prosperity and social equity, to the benefit of current and future generations (Kirchherr et al., 2017).

Mining is a global industry, and it is often located in remote, ecologically sensitive, and less-developed areas that include indigenous land and territories. If managed poorly, mining can lead to environmental degradations, displaced populations, inequality, and increased conflict, among other challenges (World Economic Forum, 2016). In the report “Mapping Mining to the Sustainable Development Goals: An Atlas, 17 Sustainable Development Goals (SDGs) are presented. This report presents an overview of the opportunities and challenges that the mining industry needs to address, from exploration through production and finally mine closure. SDG6 is about water in mining, the goal includes reduce water consumption, use alternatives water sources as seawater and greywater, integrate technical, social, economic, and political water concerns and identify high-value water areas. About waste, SDG 12 is about mining and responsible consumption, the mining industry should aim to minimize the production of waste and give value to waste (World Economic Forum, 2016).

2.2. Water supply for mining Mining consumes large quantities of water; in Chile, water consumption in mining was 62.7 m3/s in 2018, where 72% correspond to recirculated water, 22% to freshwater from inland sources, and 6% to seawater (COCHILCO, 2019a). However, mine water use accounts for a small portion of overall use (Gunson et al., 2012). Nevertheless, when mining sites are located in areas where water is a scarce resource, then mine water consumption can have a significant impact on local water resources (Northey et al., 2016). There is clear overexploitation of water resources in northern Chile, water resources are insufficient to meet environmental, domestic, and industrial requirements (Aitken et al., 2016).

Most mineral processing plants use wet processes, such as flotation, which typically takes place at 25-30% solids by mass (Gunson et al., 2012). Water and energy are interconnected as water needs energy to be treated and transported until the processing plant. The mining industry consumes vast quantities of energy. In Chile, copper mining consumed 176,745 TJ in 2018, which represents 14% of overall energy consumption in the country (COCHILCO, 2019b). As freshwater sources are getting scarce, water from other sources is needed. Recycled water from the processes and desalinated water are options to fulfill the water requirements in the processing plant. Seawater requires energy to be desalinated and transported until the mining site (Cisternas and Gálvez, 2018). Nowadays, freshwater sources are scarce, hence other water sources are needed to fulfill the demand for water that the mining industry requires. The advances in desalination technologies, reduction of membrane cost, and lower energy consumption have made seawater a viable source of freshwater for mining (Knops et al., 2013). Reverse osmosis (RO) is the desalination process most utilized worldwide due to its simplicity and lower energy consumption in comparison to distillation- based thermal processes (Shenvi et al., 2015). 2.2.1. Optimization of water supply systems The water supply system consists of the water treatment plant, pipeline, and pump station to supply water to an industrial site or urban community. Methodologies to design water supply systems usually result in mixed-integer non-linear problems (MINLP) models (Araya et al., 2018). The water supply system is transformed into a network to be analyzed, and this is called Water Distribution Network (WDN). Optimization of WDN has been done by several authors with different approaches (Ahmetović and Grossmann, 2011; Atilhan et al., 2011; El‐Halwagi, 1992; Herrera- León et al., 2019; Lira-Barragán et al., 2016; Liu et al., 2011). An integrated water supply system has demonstrated to be more efficient economically and technically than having an independent water supply to each mining processing plant (Herrera-León et al., 2019).

The reliability of WDN can be improved by adding a parallel pipeline (Gupta et al., 2014). Parallel pipelines are arranged to divide flow, keeping the head loss at the same rate, meaning that head loss in all parallel pipes remains the same (Swamee, 2001). 2.2.2. Energy production in water supply systems The use of pumps as turbines (PATS) could be an option to produce energy in the network that can be used to reduce the dependency of external sources of energy. PATs have been used in WDN instead of pressure reducing valves to reduce pressure. Additionally, PATs can produce electricity the same way turbines are used in hydropower generation (Corcoran et al., 2016; De Marchis et al., 2014; McNabola et al., 2014; Tricarico et al., 2014; Zakkour et al., 2002). The energy efficiency of water supply systems can be increased through the recovery of hydraulic energy implicit in the volumes of water transported through the supply system (Nogueira-Vilanova and Perrella-Balestieri, 2014b). 2.3. Mine tailings A critical issue in mining is the waste streams that are discarded from the processing plants, being mine tailings, the main waste produced in mineral processing plants, obtained after using wet processes, such as flotation. Mine tailings consist in a combination of fine-grained solid materials remaining after the recoverable metals and minerals have been extracted from mined ore, together with the water used in the recovery process (Industry.gov.au, 2016), which includes dissolved metals and ore processing reagents (Edraki et al., 2014). Mine tailings can account for 95-99% of the crushed and ground ores (Edraki et al., 2014). 2.3.1. Environmental impacts of mine tailings Environmental implications of tailings dam failures include human fatalities, contamination of water, and soil (Kossoff et al., 2014). One of the most severe environmental problems associated with mine tailings is acid mine drainage, which is produced when sulfide minerals are exposed to oxygen and water (Akcil and Koldas, 2006). Acid mine drainage is one of the most significant forms of water pollution (Moodley et al., 2017). Mine tailings management is a crucial issue in mining operations because of the irreversible impacts of tailings (Adiansyah et al., 2015). 2.3.2. Water reclamation in mine tailings There are a variety of ways to stored mine tailings. Conventional tailing disposal consists of transport tailings as a slurry to an impoundment or dam. Large volumes are required to be transported with the tailings to a storage facility (Edraki et al., 2014). Tailings in conventional residues disposal have approximately 60% water content. Dewatering technologies, such as thickeners and filters, can reduce water content in tailings until 20% of water (Gunson et al., 2012). Water-energy nexus in mine tailings is associated with the pumping system and the technology used in processing the tailings (Adiansyah et al., 2016). Environmental impacts and land use aspects should be considered when choosing a tailing disposal method (Adiansyah et al., 2017).

2.3.3. Re-processing of mine tailings Re-processing of tailings with the objective of extract valuable elements is a novel approach of mine tailing management. Studies about geochemical content of mine tailings have been done with samples of tailings storage facilities to analyze the content of valuable elements such as critical materials (Ceniceros-Gómez et al., 2018; Markovaara-Koivisto et al., 2018; Moran-Palacios et al., 2019; Tunsu et al., 2019). The next step seems to be the re-processing of secondary sources to obtain elements of interest. However, the environmental implications are still unknown, as re-processing tailing with such purpose is still not at an industrial level. Further analysis of Reuse opportunities of mine tailings is needed in the ongoing discussion of sustainable tailings management (Adiansyah et al., 2017).

Critical raw materials (CRMs) are those raw materials that are highly important to the European economy and but have high risk associated with their supply (European Commission, 2017). The most recent list of CRMs includes 27 raw materials, which consists of three groups elements: heavy rare earth elements (HREEs), light rare earth elements (LREEs), and PGMs (platinum group metals). As the supply of these materials is of high risk, the diversification of their sources is needed. Secondary sources like recycling from landfills and industrials waste as mine tailings appear as options for obtaining CRMs. Re-mining wastes and re-processing tailings are possible connections between the primary resource sector and the waste management sector, which would close a loop in the circular economy (Lèbre et al., 2017).

The potential use of secondary sources, like landfills and mine tailings facilities, faces several challenges. The environmental impacts will vary according to the characteristics of the minerals and rocks and techniques used. There is a need for statistics about waste facilities with official data about volumes and composition of waste sites (Careddu et al., 2018). The valorization of mine tailings is still at its infancy. To valorize mine tailings, the mining industry needs to advance in filling the knowledge gaps of mine tailings management. The situation is expected to improve as the mining industry needs to shift from a linear economy to a circular economy (Kinnunen and Kaksonen, 2019).

3. Research Methodology After having collected knowledge from literature, methodologies were designed to examine the issues identified in the previous stage. Next, the required quantitative data were collected from research articles, public government reports and datasheets, and company reports. The research methodology is divided into two sections. Section 1 explains the methodology developed to fulfill objective 1, and Section 2 describes the methodology developed to achieve objectives 2 and 3. 3.1. Water Distribution Network (WDN) design for areas with complex topography considering the energy recovery devices This section is intended to explain the process and development of the methodology designed to answer objective 1 and that is later presented in this document in publication “Design of Desalinated Water Distribution Networks: Complex Topography, Energy Production, and Parallel Pipelines.” 3.1.1. Optimization of Water Distribution Networks (WDN) Water supply systems are usually divided into four components: water intake, storage and water treatment, water pumping, and the water pipelines for the water distribution. The WDN for this thesis will be defined as the set of water treatment plants and the water distribution network, which is composed of pumping stations and the pipeline to a mining plant.

To analyze WDN, basic principles of fluid mechanics have applied: the continuity equation, the momentum principle (or conservation of momentum), and the energy equation. A related principle is the Bernoulli equation, which derives from the motion equation (Chanson, 2004). Darcy-Weisbach and Hazen Williams equations are applied to know head loss along pipelines (Swamee, 2001).

WDN design is a complex problem that includes the determination of demand, the optimal layout of WDN, the dimension of the components like pipes pumps, valves, tanks, and other components (Amit and Ramachandran, 2009). When designing a WDN, the goal is to reduce the initial investment and operational costs of the system. The primary constraint is that the desired demands are supplied with adequate pressure heads at withdrawal locations. Additionally, the water flow in a WDN, and the pressure heads at nodes must satisfy the governing laws of conservation of mass and energy (Amit and Ramachandran, 2009). Providing water at minimal cost results in an optimization problem, which has been solved using linear programming (LP), non-linear programming (NLP), Mixed Integer Non-Linear Programming (MINLP), Mixed Integer Linear Programming (MILP), dynamic programming, and heuristic- based optimization methods (Amit and Ramachandran, 2009; D’Ambrosio et al., 2015). 3.1.2. Problem Statement To achieve objective number 1, a methodology was designed to represent the WDN, including the selection of desalination plants, specifically RO plants, to supply mining

18 sites. The model includes energy recovery devices for energy production located in areas where due to the difference of altitude between 2 points is possible to produce energy like in the hydropower industry, these energy recovery devices are known as PATs. Additionally, parallel pipelines are considered to split the water flow when needed.

Mine sites are located away from the coast; these sites have a water requirement that needs to be fulfilled with desalinated water coming from RO plants located on the coast. The area features a complex topography, which means several changes in altitude from the coast until the mine site. The only option to fulfill the water requirement needed is desalinated water from RO plants. The methodology proposed by Herrera-León et al., 2019 was used as a reference for this methodology. 3.1.3. Mathematical Formulation A superstructure was used to represent the WDN, which includes RO plants, pipelines, pump stations, PATs, and the mine sites as nodes. Distances between nodes and altitude of each node were designed base on topography data obtained in Google Earth. The superstructure also includes the option to use parallel pipelines.

The objective function minimized the total annualized cost of RO plants, pipes, pump stations, and PATs required to supply several mine sites located in an isolated area far away from the coast. The objective function is not linear, so the problem results in an MINLP problem, which is transformed into a MILP problem using a piecewise methodology extracted from Lin et al., 2013. This methodology allows converting a non-linear programming problem into a linear programming problem or a mixed- integer convex programming problem for obtaining an approximated global optimum solution. 3.2. An economic evaluation of re-processing of mine tailings to obtain critical materials This section is intended to explain the methodology designed to complete objectives 2 and 3. It also presents an overview of the methods applied in publications 2 and 3. 3.2.1. Project valuation Tools Project valuation is probably the most crucial part of the selection process because it assigns a monetary value to the project. Broadly defined, the project value is the net difference between the project revenues and costs over its entire life cycle. If the net revenues of the project during the production phase are higher than the investment costs, the project is considered worthy of investment (Kodukula and Papudesu, 2006). The quality of the project valuation will depend on how they effectively include these three factors:

 Cash flow streams through the entire life cycle of the project  The discount rate used to discount future cash flows to account for their uncertainty

 Availability of management’s contingent decisions to change the course of the project Project cash flows represent the revenues and the costs of the project through the entire life cycle of the project. The costs should include the initial investment for the project and the production phase. Valuation methods assume that money today is worth more than in the future, so to acknowledge that a discount rate is used. The discount rate is the rate that is utilized to convert the future value of the project cash flows to today’s money. The discount rate is adjusted to the risk perceived to be associated with the project, so the higher the risk, the higher the discount rate (Kodukula and Papudesu, 2006). 3.2.2. Discounted Cash Flow (DCF) Discounted Cash Flow (DCF) is a valuation method that is based on calculating future cash flows of a project to decide whether it is feasible or not to invest. DCF is based on the estimation of the Net Present Value (NPV) of the project in its entire project life (Kodukula and Papudesu, 2006).

NPV is the preferred metric for project valuation under most circumstances (Arnold, 2014). To calculate NPV, cash outflows, which are the costs of the project, are deducted from the cash inflows, which represents the revenues of the project. Additionally, all cash flows are discounted to generate NPV. If the NPV is positive, then the project is feasible because the revenues exceed the expenses of the project. On the contrary, if the NPV is negative, then the project is not feasible because the costs are higher than the revenues. If the NPV is zero, then the return of the project is equal to the expenses.

Traditional valuation tools, including the DCF method, have been employed for many decades in the economic evaluation of projects. Although these methods have been effective in many cases, under specific conditions, they have certain challenges. One of the biggest dilemmas when using DCF is to choose a discount rate. The main factors that determine the discount rate for a given cash flow stream is the magnitude and the type of risk (Kodukula and Papudesu, 2006). An important consideration to determine the discount rate is whether there is uncertainty associated with the cash flow streams.

There are several methods designed to consider uncertainty when calculating the NPV, such as: to increase the discount rate, to apply sensitivity analysis, to compare pessimistic and optimistic cash flows or, to estimate the expected cash flows through scenario planning and the probability distribution (Gaspars-Wieloch, 2019). 3.2.3. Real Options Analysis (ROA) When there is considerable uncertainty related to the project cash flows, and contingent decisions are involved, traditional tools don’t include the flexibility to change the course of the project and its possible outcome (Kodukula and Papudesu, 2006). ROA addresses the issue of choices a manager may have throughout the life of the project and how those choices can enhance the value of a project (Arnold, 2014).

ROA is a complement to traditional methods as the DCF method, which seeks to analyze different possible choices for an investment project to give flexibility to the project, unlike conventional valuation methods. A real option is a right, but not an obligation, to undertake business initiatives that are connected to and exist on real assets or within real assets (Trigeorgis, 1993).

ROA adds value to an economic analysis like the DCF method by considering different options such as wait to invest, abandon an investment, or expand a business. Then, more opportunities and flexibility are available when a project is surrounded by considerable uncertainty. ROA can be applied using different methods such as Black & Scholes Equation, Decision trees or binomial trees, Datar-Mathews model, which uses cash flow scenarios combined with Monte Carlo simulation, and Fuzzy Pay-off Method (Collan, 2011). 3.2.4. Sensitivity Analysis As a complement to the DCF method, sensitivity analysis (SA) can be performed on the NPV to explore the sensitivity of key input parameters. Sensitivity analysis is the study of how the uncertainty in the inputs of a mathematical model or system can be distributed into different sources of uncertainty in its inputs. Alongside sensitivity analysis, uncertainty analysis is usually performed. Uncertainty analysis assesses the uncertainty in model outputs that derives from the uncertainty in the inputs. Methods to perform SA include simulation and scenario techniques. Global sensitivity analysis methods consider the changes in model outputs as input factors change all together over specified ranges. These methods have the ability to manage non-linear and non- additive responses and to observe relationships between multiple factors.

Monte Carlo Simulation is a method that consists of the simulation of thousands of possible project scenarios, calculation of the NPV for each scenario using the DFC method, and analyzing the probability distribution of the NPV results (Kodukula and Papudesu, 2006).

4. Publications and Results Reviews An overview of the publications is included in chapter 4, which includes the objectives, findings, and contributions of each publication of this doctoral dissertation. The overall purpose of the dissertation is to… 4.1. Publication I: Design of desalinated water distribution networks: Complex topography, energy production, and parallel pipelines 4.1.1. Research Objective The objective of this publication is to provide a methodology to design the water distribution network (WDN) and reverse osmosis (RO) plants to supply mining companies with desalinated water, the mathematical model considers the following elements to design the WDN: energy recovery devices, parallel pipelines, and complex topography. Energy recovery devices considered in the model are PATs for producing energy in the network using hydropower principles. Parallel pipelines are considered when the water requirement is large and cannot be supplied with one pipeline. In this study complex topography means 4.1.2. Contributions PATs are often used in urban water distribution networks to reduce pressure, and there are some applications to produce energy. The main contribution of this study is proposing the use of energy recovery devices in areas with the mining industry where they are yet not used. These devices can be used in areas with complex topography, meaning isolated areas and presents changes in elevation as the pipelines go from the seashore to the mining site. Mining sites are usually located in areas suffering from water scarcity; thus, the use of desalinated water is a must.

A methodology to simultaneously design the WDN to supply several mining sites located in an area with complex topography, additionally to find the location and sizes of desalination plants, pipelines, pump stations, and energy recovery devices, was addressed. A case study based on an area in northern Chile was used to validate the model. A superstructure was used to represent the WDN, which includes RO plants, pipelines, pump stations, and energy recovery devices. The objective of the model is to minimize the total annual cost by finding the optimal WDN.

Results show the use of PATs in a region with complex topography is indeed feasible for producing energy in the WDN in areas like the Coastal Cordillera, where larger elevations in altitude are often avoided by mining companies. The use of parallel pipelines is a feasible option to supply with desalinated water to several mining sites, instead of having their own RO plant and own pipeline system for each mining site.

4.2. Publication II: Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings 4.2.1. Research Objective This study aims to conduct a technical and economic assessment of the valorization of mine tailings of Chile as a source of critical materials. In recent years, the use of secondary sources, such as mining waste and electrical waste, has gained importance. This research adds knowledge to that field of study by adding a novel techno- economic assessment for producing critical materials by re-processing mine tailings. 4.2.2. Contributions Results show that mine tailings facilities of the copper industry in Chile store valuable elements such as critical materials. Therefore, the evaluation of geochemical content alongside the identification of suitable technologies, and an economic analysis will help to find alternative sources of critical materials, as is the case of re-processing mine tailings with such purpose.

The DCF method is a method widely used in economic assessment but is not a decisive metric for a final decision on real investment. To ensure the robustness of the assessment, a sensitivity analysis was performed on the results of the NPV, by analyzing the effect of the market prices of critical materials, capital, and operating costs on the options assessed for producing critical materials. The options analyzed were producing rare earths concentrate or producing vanadium pentoxide. Results show that under certain conditions the production of vanadium pentoxide is feasible.

The main contribution of this study is to show the economic potential of Chilean mine tailings by performing a techno-economic assessment of the valorization of tailings. 4.3. Publication III: Feasibility of re-processing mine tailings to obtain critical materials using real options analysis (submitted in Sustainable Development) 4.3.1. Research objective This study aims to propose a framework to assess the feasibility of re-processing mine tailings to obtain critical materials, considering flexibility aspects, such as waiting for investment in the case of a negative outcome in the present.

The NPV estimated with the DCF is used as a starting point; then, with ROA, different outcomes for an investment project are assessed. ROA is applied using binomial tree analysis. The binomial tree is built by using the method of risk-neutral probabilities.

For re-processing mine tailings, technology development and business models are still needed. To study the influence of several variables on the NPV outcome, Monte Carlo simulation is applied to perform sensitivity and uncertainty analyses.

The framework developed in this study is applied to a case study based on active mine tailings located in the Antofagasta Region in Northern Chile. This region features

23 several mining projects based on copper resources, leaving tremendous volumes of mine tailings that are stored without any economic purpose. 4.3.2. Contributions The novelty of this study is to consider ROA and sensitivity analysis to provide flexibility to the economic assessment of re-processing mine tailings, acknowledging the uncertainties involved. Traditional valuation tools such as the DCF method are statistic methods that do not consider the uncertainty of the variables used to estimate the profitability of an investment.

ROA gives additional value, so decision-makers can explore alternatives while waiting to uncertainty to clear off to invest, to re-estimate the project payoff, or they can decide to abandon a project.

5. Conclusions Reducing primary ore extraction and reducing waste are part of the strategies that the mining industry needs to apply for implementing sustainable manufacturing. Reducing approaches in every manufacturing and processing industry is the core of the circular economy. Re-processing mine tailings will reduce the amount of waste that a mining plant left behind, and it will reduce the need for primary mining ores. Another challenge for the mining industry is to reduce and optimize the demand for freshwater, especially in arid zones where water is a scarce resource. The purpose of this thesis is to develop methodologies to improve water and mine tailings management in mining processes.

The conclusions, related to the research questions, are presented below: RQ 1.1. Is an integrated water supply system that includes energy recovery devices a feasible option in areas with complex topography?

A methodology was developed to simultaneously design a water distribution network to supply mining plants located in an area with complex topography with desalinated water, as well as the location and size of desalination plants, pipelines, pump stations, and PATs stations. The methodology was validated with a case study that corresponds to a geographic area located in the Antofagasta region, which includes four mining plants (Araya et al., 2018). This area features several changes in altitude, as is located in the Atacama Desert, which is cross by the Coastal Cordillera.

The optimal solution resulted in a WDN that considers one desalination plant to supply the four mining plants, as well as PATs located in locations with altitude. Additionally, the optimal solution included the use of parallel pipelines. A ranking of solutions was presented, which showed that an integrated system prevails to the traditional water supply system, which consists of a desalination plant and a set of pumping stations and pipelines to each mining plant (Araya et al., 2018). RQ 1.2. Is it economically and technically feasible to use energy recovery devices in the water supply system?

The use of energy recovery devices, in this case, PATs is feasible in the presence of a great difference in altitude between two nodes. Having an integrated system to supply desalinated water to several mining plants reduces the costs of the water supply system and enhances the industrial symbiosis between different companies. Moreover, the use of PATs can reduce the costs by producing energy, which can be used in the water distribution network (Araya et al., 2018, 2017).

The difference of altitude between two nodes, in which the option of using PATS was considered, was varied to understand in which situations PATs are a feasible option, from both technical and economic aspects. This analysis showed that there is a difference in altitude where the option of PATs is no technically possible(Araya et al., 2018).

RQ 1.3. When using parallel pipes is a feasible option in the water supply system?

Parallel pipes are needed when the water flow is larger than the maximum water flow allowed in industrial pipelines, so in the case of an integrated water supply system using parallel pipes is a feasible option to supply water to several mining companies. The optimal solution features parallel pipes, as a set of three pipes and two pipes. Furthermore, the three first solutions include parallel pipes, demonstrating that parallel pipes are needed when having to supply several mines (Araya et al., 2018). RQ 2.1. What critical materials can be recovered from mine tailings?

Mine tailings contain several elements, including some critical materials. A methodology to assess the valorization of mine tailings was developed, with the focus on the recovery of critical materials. The methodology was validated with a case study that corresponds to inactive mine tailings deposits located in the Antofagasta region (Araya et al., 2020). Mine tailings of copper mining in Chile contains several critical materials, including rare earths, vanadium, cobalt, and scandium. A ranking of critical materials that could be extracted from mine tailings was presented, the critical materials were evaluated according to the quantity present in tailings deposits and the price of the critical material. This analysis showed that rare earths, cobalt, scandium, and vanadium could be extracted from inactive mine tailings in the Antofagasta Region, due to their price and quantity (Araya et al., 2020). RQ 2.2. What are the challenges in the production of critical materials using mine tailings as a source?

Emerging technologies for recovering critical materials from mine tailings were reviewed, showing that Technology development is still much needed to make the recovery of critical materials from mine tailings at an industrial scale. At this time, most of the research is still in the laboratory and pilot-scale (Araya et al., 2020). Further research should address new strategies to anticipate the future use of material beyond the closing of a mine (Lèbre et al., 2017). RQ 2.3. Is it economically feasible to invest in a project developed around the idea of re- processing mine tailings to obtain critical materials?

The economic assessment developed to assess the production of critical materials from mine tailings showed that, in certain conditions, the recovery of critical materials is feasible. DCF method was used to calculate the NPV of two projects, one based on the production of rare earths concentrate and another one based on the production of vanadium pentoxide. Results showed that a project based on the recovery of vanadium pentoxide is feasible (Araya et al., 2020).

Alongside the DCF method, a sensitivity analysis was performed on the inputs of NPV, to understand which variables have a more significant effect on the NPV. The inputs variables studied were capital expenditures, operational expenditures, price of the critical material, and discount rate (Araya et al., 2020).

RQ 3.1. How can real options improve the decision to invest in a project using mine tailings as a source of critical materials?

A project based on the recovery of critical materials from mine tailings can have great uncertainty in their inputs. ROA complements traditional valuation methods such as the DCF, by including other options for investment. ROA provides flexibility to a project based on the re-processing of mine tailings, acknowledging the uncertainties involved. RQ 3.2. How can the net present value variables influence real options performance?

The inputs of the NPV can have uncertainties that affect the outcome. Using the NPV as a decisive metric for investments can lead to errors in the decision-making process. The NPV is used as a starting point to ROA since it is a complement to traditional tools rather than a method by itself. 5.1.Theoretical Contribution of the Study This thesis contributes with a collection of methodologies to assess water management and mine tailings management in mining. The mining industry is facing many challenges as industries need to fulfill sustainable development goals to achieve a circular economy.

The first contribution of this thesis is to provide a methodology to design integrated water supply systems to provide desalinated water to several mining plants. The novel contribution of this methodology is to include the option of placing energy recovery devices in locations where the difference of altitude can be advantageous to produce energy as it is done in hydropower plants. Additionally, the methodology includes parallel pipes to fulfill the demand of several mine plants at the same time. The other contributions of the thesis are embedded in the field of mine tailings management. These contributions are a methodology to assess the feasibility of re- processing mine tailings to obtain critical materials and a methodology to assess the feasibility of re-processing mine tailings with a real options approach. The novelty of theses methodologies is to assess the feasibility of projects based on the idea of obtaining critical materials from a secondary source such as mine tailings. 5.2.Limitation and Future Research Suggestions A limitation of this study is the difficulty to find data about technologies suitable for recovering critical materials from mine tailings. Data about capital expenses, operational expenses, and prices are not easy to find.

Another limitation is that all the methodologies presented in this thesis are validated with case studies of the copper mining industry in the Antofagasta Region. These methodologies could be applied to other case studies, but some considerations and changes should be made.

Future research suggestions will be focused on implementing the circular economy in mining processes with a focus on mine tailings management. Mine tailings are the 27 biggest sink of water of the mineral processing, so efforts should be made to create a framework that integrates water management for mine tailings with the re-processing of mine tailings. Future studies should also incorporate an environmental assessment of projects based on the re-processing of mine tailings.

References

Adiansyah, J.S., Haque, N., Rosano, M., Biswas, W., 2017. Application of a life cycle assessment to compare environmental performance in coal mine tailings management. J. Environ. Manage. 199, 181–191. https://doi.org/10.1016/j.jenvman.2017.05.050 Adiansyah, J.S., Rosano, M., Vink, S., Keir, G., 2015. A framework for a sustainable approach to mine tailings management: Disposal strategies. J. Clean. Prod. 108, 1050–1062. https://doi.org/10.1016/j.jclepro.2015.07.139 Adiansyah, J.S., Rosano, M., Vink, S., Keir, G., Stokes, J.R., 2016. Synergising water and energy requirements to improve sustainability performance in mine tailings management. J. Clean. Prod. 133, 5–17. https://doi.org/10.1016/j.jclepro.2016.05.100 Adibi, N., Ataee-Pour, M., Rahmanpour, M., 2015. Integration of sustainable development concepts in open pit mine design. J. Clean. Prod. 108, 1037–1049. https://doi.org/10.1016/j.jclepro.2015.07.150 Ahmetović, E., Grossmann, I.E., 2011. Global superstructure optimization for the design of integrated process water networks. AIChE J. 57, 434–457. https://doi.org/10.1002/aic.12276 Aitken, D., Rivera, D., Godoy-Faúndez, A., Holzapfel, E., 2016. Water scarcity and the impact of the mining and agricultural sectors in Chile. Sustain. 8. https://doi.org/10.3390/su8020128 Akcil, A., Koldas, S., 2006. Acid Mine Drainage (AMD): causes, treatment and case studies. J. Clean. Prod. 14, 1139–1145. https://doi.org/10.1016/j.jclepro.2004.09.006 Amit, R.K., Ramachandran, P., 2009. Optimal Design of Water Distribution Networks A Review. Araya, N., Cisternas, L.A., Lucay, F., Gálvez, E.D., 2017. Design of desalinated water distribution networks including energy recovery devices, Computer Aided Chemical Engineering. https://doi.org/10.1016/B978-0-444-63965-3.50156-2 Araya, N., Kraslawski, A., Cisternas, L.A., 2020. Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings. J. Clean. Prod. 263, 121555. https://doi.org/10.1016/j.jclepro.2020.121555 Araya, N., Lucay, F.A., Cisternas, L.A., Gálvez, E.D., 2018. Design of Desalinated Water Distribution Networks: Complex Topography, Energy Production, and Parallel Pipelines. Ind. Eng. Chem. Res. 57. https://doi.org/10.1021/acs.iecr.7b05247 Arnold, T., 2014. A Pragmatic Guide to Real Options, A Pragmatic Guide to Real Options. Palgrave Macmillan, New York.

https://doi.org/10.1057/9781137391162 Atilhan, S., Linke, P., Wahab, A.A., Halwagi, M.M. El, 2011. A systems integration approach to the design of regional water desalination and supply networks. Int. J. Process Syst. Eng. 1, 125. https://doi.org/10.1504/ijpse.2011.038941 Binnemans, K., Jones, P.T., Blanpain, B., Van Gerven, T., Pontikes, Y., 2015. Towards zero-waste valorisation of rare-earth-containing industrial process residues: A critical review. J. Clean. Prod. 99, 17–38. https://doi.org/10.1016/j.jclepro.2015.02.089 Careddu, N., Dino, G.A., Danielsen, S.W., Přikryl, R., 2018. Raw materials associated with extractive industry: An overview. Resour. Policy 59, 1–6. https://doi.org/10.1016/j.resourpol.2018.09.014 Caron, J., Durand, S., Asselin, H., 2016. Principles and criteria of sustainable development for the mineral exploration industry. J. Clean. Prod. 119, 215–222. https://doi.org/10.1016/j.jclepro.2016.01.073 Ceniceros-Gómez, A.E., Macías-Macías, K.Y., de la Cruz-Moreno, J.E., Gutiérrez-Ruiz, M.E., Martínez-Jardines, L.G., 2018. Characterization of mining tailings in México for the possible recovery of strategic elements. J. South Am. Earth Sci. 88, 72–79. https://doi.org/10.1016/j.jsames.2018.08.013 Chanson, H., 2004. Hydraulics of Open Channel Flow, Second Edi. ed. https://doi.org/https://doi.org/10.1016/B978-0-7506-5978-9.X5000-4 Cisternas, L.A., Gálvez, E.D., 2018. The use of seawater in mining. Miner. Process. Extr. Metall. Rev. 39, 18–33. https://doi.org/10.1080/08827508.2017.1389729 COCHILCO, 2019a. Consumo de agua en la minería del cobre al 2018 - Comisión Chilena del Cobre [WWW Document]. URL https://www.cochilco.cl/Listado Temtico/Consumo de agua en la mineria del cobre al 2018 - Version Final.pdf (accessed 1.1.20). COCHILCO, 2019b. Informe de actualización del consumo energético de la minería del cobre al año 2018. Com. Chil. del Cobre, Minist. Minería 31. https://doi.org/DE 11/2016 Collan, M., 2011. Thoughts about selected models for the valuation of real options. Acta Univ. Palacki. Olomuc. Fac. Rerum Nat. Math. 50, 5–12. Corcoran, L., McNabola, A., Coughlan, P., 2016. Optimization of water distribution networks for combined hydropower energy recovery and leakage reduction. J. Water Resour. Plan. Manag. 142, 1–8. https://doi.org/10.1061/(ASCE)WR.1943- 5452.0000566 D’Ambrosio, C., Lodi, A., Wiese, S., Bragalli, C., 2015. Mathematical programming techniques in water network optimization. Eur. J. Oper. Res. 243, 774–788. https://doi.org/10.1016/j.ejor.2014.12.039 De Marchis, M., Fontanazz, C.M., Freni, G., Messineo, A., Milici, B., Napoli, E., Notaro,

V., Puleo, V., Scopa, A., 2014. Energy recovery in water distribution networks. Implementation of pumps as turbine in a dynamic numerical model. Procedia Eng. 70, 439–448. https://doi.org/10.1016/j.proeng.2014.02.049 Dong, L., Tong, X., Li, X., Zhou, J., Wang, S., Liu, B., 2019. Some developments and new insights of environmental problems and deep mining strategy for cleaner production in mines. J. Clean. Prod. 210, 1562–1578. https://doi.org/10.1016/j.jclepro.2018.10.291 Edraki, M., Baumgartl, T., Manlapig, E., Bradshaw, D., Franks, D.M., Moran, C.J., 2014. Designing mine tailings for better environmental, social and economic outcomes: A review of alternative approaches. J. Clean. Prod. 84, 411–420. https://doi.org/10.1016/j.jclepro.2014.04.079 El‐Halwagi, M.M., 1992. Synthesis of reverse‐osmosis networks for waste reduction. AIChE J. 38, 1185–1198. https://doi.org/10.1002/aic.690380806 Elshkaki, A., Graedel, T.E., Ciacci, L., Reck, B., 2016. Copper demand, supply, and associated energy use to 2050. Glob. Environ. Chang. 39, 305–315. https://doi.org/10.1016/j.gloenvcha.2016.06.006 European Commission, 2017. Critical raw materials | Internal Market, Industry, Entrepreneurship and SMEs [WWW Document]. URL https://ec.europa.eu/growth/sectors/raw-materials/specific- interest/critical_en (accessed 5.19.20). Falagán, C., Grail, B.M., Johnson, D.B., 2017. New approaches for extracting and recovering metals from mine tailings. Miner. Eng. 106, 71–78. https://doi.org/10.1016/j.mineng.2016.10.008 Gaspars-Wieloch, H., 2019. Project Net Present Value estimation under uncertainty. Cent. Eur. J. Oper. Res. 27, 179–197. https://doi.org/10.1007/s10100-017-0500-0 Ghorbani, Y., Kuan, S.H., 2017. A review of sustainable development in the Chilean mining sector: past, present and future. Int. J. Mining, Reclam. Environ. 31, 137– 165. https://doi.org/10.1080/17480930.2015.1128799 Gorman, M.R., Dzombak, D.A., 2018. A review of sustainable mining and resource management: Transitioning from the life cycle of the mine to the life cycle of the mineral. Resour. Conserv. Recycl. 137, 281–291. https://doi.org/10.1016/j.resconrec.2018.06.001 Gunson, A.J., Klein, B., Veiga, M., Dunbar, S., 2012. Reducing mine water requirements. J. Clean. Prod. 21, 71–82. https://doi.org/10.1016/j.jclepro.2011.08.020 Gunson, A.J., Klein, B., Veiga, M., Dunbar, S., 2010. Reducing mine water network energy requirements. J. Clean. Prod. 18, 1328–1338. https://doi.org/10.1016/j.jclepro.2010.04.002 Gupta, R., Baby, A., Arya, P. V., Ormsbee, L., 2014. Upgrading reliability of water distribution networks recognizing valve locations. Procedia Eng. 89, 370–377.

https://doi.org/10.1016/j.proeng.2014.11.201 Herrera-León, S., Lucay, F.A., Cisternas, L.A., Kraslawski, A., 2019. Applying a multiobjective optimization approach in designing water supply systems for mining industries. The case of Chile. J. Clean. Prod. 210. https://doi.org/10.1016/j.jclepro.2018.11.081 Ihle, C.F., Kracht, W., 2018. The relevance of water recirculation in large scale mineral processing plants with a remote water supply. J. Clean. Prod. 177, 34–51. https://doi.org/10.1016/j.jclepro.2017.12.219 Industry.gov.au, 2016. Tailings management - Leading Practice Sustainable Development Program for the Mining I, Australian Government. Kinnunen, P.H.M., Kaksonen, A.H., 2019. Towards circular economy in mining: Opportunities and bottlenecks for tailings valorization. J. Clean. Prod. 228, 153– 160. https://doi.org/10.1016/j.jclepro.2019.04.171 Kirchherr, J., Reike, D., Hekkert, M., 2017. Conceptualizing the circular economy: An analysis of 114 definitions. Resour. Conserv. Recycl. 127, 221–232. https://doi.org/10.1016/j.resconrec.2017.09.005 Knops, F., Kahne, E., de la Mata, M.G., Fajardo, C.M., 2013. Seawater desalination off the Chilean coast for water supply to the mining industry. Desalin. Water Treat. 51, 11–18. https://doi.org/10.1080/19443994.2012.694233 Kodukula, P., Papudesu, C., 2006. Project Valuation Using Real Options - A practitioner’s guide. J. Ross Publishing, Fort Lauderdale, Florida. Kossoff, D., Dubbin, W.E., Alfredsson, M., Edwards, S.J., Macklin, M.G., Hudson- Edwards, K.A., 2014. Mine tailings dams: Characteristics, failure, environmental impacts, and remediation. Appl. Geochemistry 51, 229–245. https://doi.org/10.1016/j.apgeochem.2014.09.010 Lèbre, É., Corder, G., Golev, A., 2017. The Role of the Mining Industry in a Circular Economy: A Framework for Resource Management at the Mine Site Level. J. Ind. Ecol. 21, 662–672. https://doi.org/10.1111/jiec.12596 Lin, M.H., Carlsson, J.G., Ge, D., Shi, J., Tsai, J.F., 2013. A review of piecewise linearization methods. Math. Probl. Eng. 2013. https://doi.org/10.1155/2013/101376 Lira-Barragán, L.F., Ponce-Ortega, J.M., Guillén-Gosálbez, G., El-Halwagi, M.M., 2016. Optimal Water Management under Uncertainty for Shale Gas Production. Ind. Eng. Chem. Res. 55, 1322–1335. https://doi.org/10.1021/acs.iecr.5b02748 Liu, S., Konstantopoulou, F., Gikas, P., Papageorgiou, L.G., 2011. A mixed integer optimisation approach for integrated water resources management. Comput. Chem. Eng. 35, 858–875. https://doi.org/10.1016/j.compchemeng.2011.01.032 Lottermoser, B.G., 2011. Recycling, reuse and rehabilitation of mine wastes. Elements 7, 405–410. https://doi.org/10.2113/gselements.7.6.405

Markovaara-Koivisto, M., Valjus, T., Tarvainen, T., Huotari, T., Lerssi, J., Eklund, M., 2018. Preliminary volume and concentration estimation of the Aijala tailings pond – Evaluation of geophysical methods. Resour. Policy 59, 7–16. https://doi.org/10.1016/j.resourpol.2018.08.016 McNabola, A., Coughlan, P., Corcoran, L., Power, C., Williams, A.P., Harris, I., Gallagher, J., Styles, D., 2014. Energy recovery in the water industry using microhydropower: An opportunity to improve sustainability. Water Policy 16, 168–183. https://doi.org/10.2166/wp.2013.164 Moodley, I., Sheridan, C.M., Kappelmeyer, U., Akcil, A., 2017. Environmentally sustainable acid mine drainage remediation: Research developments with a focus on waste/by-products. Miner. Eng. 126, 1–14. https://doi.org/10.1016/j.mineng.2017.08.008 Moran-Palacios, H., Ortega-Fernandez, F., Lopez-Castaño, R., Alvarez-Cabal, J. V., 2019. The Potential of Iron Ore Tailings as Secondary Deposits of Rare Earths. Appl. Sci. 9, 2913. https://doi.org/10.3390/app9142913 Nogueira-Vilanova, M.R., Perrella-Balestieri, J.A., 2014a. Energy and hydraulic efficiency in conventional water supply systems. Renew. Sustain. Energy Rev. 30, 701–714. https://doi.org/10.1016/j.rser.2013.11.024 Nogueira-Vilanova, M.R., Perrella-Balestieri, J.A., 2014b. Hydropower recovery in water supply systems: Models and case study. Energy Convers. Manag. 84, 414– 426. https://doi.org/10.1016/j.enconman.2014.04.057 Northey, S.A., Mudd, G.M., Saarivuori, E., Wessman-Jääskeläinen, H., Haque, N., 2016. Water footprinting and mining: Where are the limitations and opportunities? J. Clean. Prod. 135, 1098–1116. https://doi.org/10.1016/j.jclepro.2016.07.024 Ramírez, Y., Kraslawski, A., Cisternas, L.A., 2019. Decision-support framework for the environmental assessment of water treatment systems. J. Clean. Prod. 225, 599– 609. https://doi.org/10.1016/j.jclepro.2019.03.319 Reporte Minero, 2018. Sector minero aporta el 10% del PIB nacional [WWW Document]. URL https://www.reporteminero.cl/noticia/noticias/2018/04/sector-minero- aporta-el-10-del-pib-nacional (accessed 6.9.20). Reyes-Bozo, L., Godoy-Faúndez, A., Herrera-Urbina, R., Higueras, P., Salazar, J.L., Valdés-González, H., Vyhmeister, E., Antizar-Ladislao, B., 2014. Greening chilean copper mining operations through industrial ecology strategies. J. Clean. Prod. 84, 671–679. https://doi.org/10.1016/j.jclepro.2014.03.088 SERNAGEOMIN, 2019. Anuario de la Minería de Chile [WWW Document]. URL https://www.sernageomin.cl/anuario-de-la-mineria-de-chile/ (accessed 10.31.19). SERNAGEOMIN, 2018. Datos Públicos Depósito de Relaves [WWW Document]. URL

https://www.sernageomin.cl/datos-publicos-deposito-de-relaves/ (accessed 5.10.18). Shenvi, S.S., Isloor, A.M., Ismail, A.F., 2015. A review on RO membrane technology: Developments and challenges. Desalination 368, 10–26. https://doi.org/10.1016/j.desal.2014.12.042 Swamee, P.K., 2001. Optimal design of pumping mains. J. Transp. Eng. 127, 264–267. https://doi.org/10.1061/(ASCE)0733-947X(2001)127:3(264) Tricarico, C., Morley, M.S., Gargano, R., Kapelan, Z., De Marinis, G., Savić, D., Granata, F., 2014. Integrated optimal cost and pressure management for water distribution systems. Procedia Eng. 70, 1659–1668. https://doi.org/10.1016/j.proeng.2014.02.183 Trigeorgis, L., 1993. The Nature of Option Interactions and the Valuation of Investments with Multiple Real Options. J. Financ. Quant. Anal. https://doi.org/10.2307/2331148 Tunsu, C., Menard, Y., Eriksen, D.Ø., Ekberg, C., Petranikova, M., 2019. Recovery of critical materials from mine tailings: A comparative study of the solvent extraction of rare earths using acidic, solvating and mixed extractant systems. J. Clean. Prod. 218, 425–437. https://doi.org/10.1016/j.jclepro.2019.01.312 World Economic Forum, 2016. Mapping Mining to the Sustainable Development Goals: An atlas [WWW Document]. URL https://www.undp.org/content/dam/undp/library/Sustainable Development/Extractives/Mapping_Mining_SDGs_An_Atlas_Executive_Sum mary_FINAL.pdf (accessed 2.7.20). Zakkour, P.D., Gaterell, M.R., Griffin, P., Gochin, R.J., Lester, J.N., 2002. Developing a sustainable energy strategy for a water utility. Part II: A review of potential technologies and approaches. J. Environ. Manage. 66, 115–125. https://doi.org/10.1006/jema.2002.0567

SECTION II: PUBLICATIONS

PUBLICATION I

Araya, N., Lucay, F.A., Cisternas, L.A., Gálvez, E.D.

Design of Desalinated Water Distribution Networks: Complex Topography, Energy Production, and Parallel Pipelines.

Industrial & Engineering Chemical Research

https://doi.org/10.1021/acs.iecr.7b05247

Vol 57 (30), 9879-9888, 2018

Cite This: Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX pubs.acs.org/IECR

Design of Desalinated Water Distribution Networks: Complex Topography, Energy Production, and Parallel Pipelines Natalia Araya,† Freddy A. Lucay,† Luis A. Cisternas,† and Edelmira D. Galveź *,‡

† Depto. de Ing. Química y Procesos de Minerales, Universidad de Antofagasta, Antofagasta, Chile ‡ Depto. de Ing. Metalurgia y Minas, Universidad Catolicá del Norte, Antofagasta, Chile

ABSTRACT: A methodology was developed to determinate the location and size of desalination plants, the water distribution network, and the location and size of energy recovery devices to provide desalinated water in regions with complex topography. The novelty of this methodology is that energy recovery devices such as pumps as turbines are incorporated to produce energy. Another novelty is the consideration of using multiple pipelines. The methodology proposed uses a superstructure with a set of alternatives in which the optimal solution is found. A mathematical model is generated that corresponds to a mixed integer nonlinear programming (MINLP) problem, which is linearized to become a mixed integer linear programming (MILP) problem solved using the CPLEX solver in GAMS. A case study is presented to demonstrate the applicability of the methodology to real size problems.

1. INTRODUCTION production (35% of the world’s copper production) and other Water and energy are vital elements to the well being of metallic elements like silver, gold, and molybdenum and nonmetallic minerals like potassium nitrate, lithium carbonate, society. At a basic level, electricity generation requires water, and boron. The Antofagasta Region is located in the Atacama and water treatment and transport require electricity. World- Desert, the driest nonpolar desert of the world, and water is wide energy consumption destined for water supply represents needed to process these ore resources. Water is a limited 7% of global energy consumption.1 Hamiche et al.2 reviewed resource due to the overexploitation of groundwater sources. the nexus between water and energy comprehensively and On the other hand, the Antofagasta Region has a large sea presented a classification system; their work suggested that this coast; therefore, seawater has become the main source of fresh nexus should be explored widely. water for the mining industry and urban populations. Nowadays, fresh water sources are scarce, and nontraditional fi The mines are often located far from the coast and, more water sources are utilized more every day to ful ll the importantly, at high altitudes. These changes in elevation are increasing demand for both human and industrial consump- the result of the presence in the coast of the Coastal Cordillera, tion. The advance in desalination processes, reduction of which is a mountain range, and the presence of the Andes See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. Downloaded via LAPPEENRANTA UNIV OF TECHNOLOGY on June 22, 2018 at 12:13:58 (UTC). membrane cost, and lower energy consumption have made 3 Mountains. Mine elevations vary between 1000 and 4000 m seawater a viable source of freshwater. above sea level. In the region, there are several mining plants Reverse osmosis (RO) is a feasible option to provide water already using desalinated water obtained through RO. Each for isolated and desert areas where there is access to seawater. mining company has a desalination plant with a water RO is the desalination process most utilized worldwide; more distribution network (WDN) to provide fresh water for its than 50% of desalination plants correspond to RO plants due processes without any integration between companies. The to its simplicity and because its energy requirements are minor ffi 4 need to develop a more e cient desalination and WDN when compared to distillation-based thermal processes. system, for example, through the integration of several WDNs, The transport of desalinated water in areas with complex motivates this work. topology, e.g., areas with mountains, is a major challenge for several reasons. On the one hand, the transport costs due to the elevations are significant, which can be several times the Special Issue: 2017 European Symposium on Computer-Aided cost of desalting, and on the other hand, finding only upward Process Engineering distribution systems is difficult. An example corresponds to the Received: December 22, 2017 north of Chile (Antofagasta Region) where there are several Revised: June 5, 2018 mining plants located at high altitude in a desert area. From Accepted: June 7, 2018 this area, Chile produces more than half of the Chilean copper Published: June 7, 2018

There have been several works regarding the design of RO model. The methodology was demonstrated with a case study, plants and water supply systems. The methodologies to design and its application was adaptable to similar problems. water supplies systems usually result in mixed integer nonlinear However, they did not consider WDN design including sites problem (MINLP) models. Amit and Ramachandran5 realized with changes in the elevations resulting from the presence of a review of the current status of optimization models for mountains. designing water distribution networks and presented recom- The use of pumps as turbines (PATs) can be an option of mendations for future research; Coehlo and Andrade-Campos1 producing energy in the network that can be used to supply provided a review about methods to achieve water supply energy for water pumping. Pumps operating as turbines can be system efficiency, which also included design optimization. used in water distribution systems to reduce pressure instead of Khor et al.6 provided state of the art water networks synthesis using pressure reducing valves. Additionally, they can produce focusing on a single site and continuous process problems, electricity.17 where major modeling and computational challenges were There are some reviews on energy efficiency in water supply discussed exploring issues such as nonconvexity and non- systems that include hydropower generation and PATs linearity. implementation. Vilanova and Balestieri18 presented models El-Halwagi7 introduced the novel notion of synthesizing RO of hydropower recovery in water supply systems that could be networks applied to waste reduction. The synthesis was applied in a water distribution network design. Zakkour et al.19 formulated as MINLP, and the objective of this work was to reviewed some emerging technologies and practices for minimize the total annualized cost of RO networks. Since then, sustainable water utility. McNabola et al.20 presented a review considerable research has been made regarding the design of of energy use in the water industry and opportunities for RO plants and water distribution systems. microhydropower (MHP) energy recovery. Liu et al.8 considered using desalinated water, wastewater, Vilanova and Balestieri21 evaluated the possibilities and and reclaimed water to supply water deficient areas using a benefits of recovering and producing energy in water supply mixed integer linear programming (MILP) model taking into systems. They illustrate technical, economic, and environ- account geographical aspects of the region; however, this work mental aspects of hydropower recovery in water supply systems did not consider using energy production. Ahmetovićand using a case study. Corcoran et al.22 developed a methodology Grossman9 proposed a general superstructure and a model to to find the optimal location of turbines in a water distribution optimize integrated process water networks using MINLP and network using a MINLP approach and an evolutionary NLP models, considering multiple sources of water and approach as a comparison. Tricarico et al.17 presented a different qualities of water; the methodology was applied in methodology to implement PATs combined with pump different case studies. Atilhan et al.10 proposed a novel scheduling to recover energy and pressure water regime in a approach for the design of desalination and water distribution water distribution network, showing clear economic benefits. networks considering first a source-interception sink repre- De Marchis et al.23 analyzed the application of PATs in a water sentation; then, an optimization problem was formulated distribution network using a hydrodynamic model. The model resulting in a NLP problem whose objective was to meet the was demonstrated to correctly represent the impact of energy requirements for the sinks at the minimum cost while satisfying recovery on water supply distribution. Carravetta et al.24 constraints for the sinks. The methodology was applied to a presented the implementation of microhydroelectric plants case study. Lira-Barragań et al.11 proposed that mathematical which included PATs in urban water networks. The project of programming models for synthetizing WDN associated with a small hydroelectric plant was performed at the inlet node of a shale gas production take into account uncertainties, and the real network. The case study showed that the installation of proposed superstructure for water integration allows the small hydroelectric plants could provide interesting economic management of fresh water consumption and wastewater benefits for the manager of pipe networks in urban areas. None streams. Liang et al.12 proposed a convex model, which of the works reviewed analyzed the implementation of PATs in corresponded to a MINLP problem, to obtain the optimal a WDN on a larger scale; furthermore, none considered design of water distribution systems using the Hanoi network, producing energy like that in the hydropower industry from which is a known problem of the looped water distribution the downfall of water in a WDN that included RO plants and network as a case study.13 Gonzalez-Bravó et al.14 proposed a industrial sites with a large requirement of water. multiobjective optimization approach for synthesizing water Parallel pipelines are arranged to divide flow and to keep the distribution networks considering domestic, agricultural, and head loss at the same rate, meaning that head loss in all parallel industrial users involving dual purpose plants considering pipes remains the same.25 The reliability of a water distribution environmental, economic, and social aspects. Gonzalez-Bravó network can be improved by adding a parallel pipe to an et al.15 presented an approach to design water and energy existing one.26 In actual industrial operations, more than one distribution networks based on a multistakeholder environ- pipeline is often utilized to provide higher flows, mainly in ment where a multiobjective model considering economic, mining operations which require a great amount of water in environmental, and social impacts was applied in a stressed their operations, so one pipeline cannot satisfy the water scheme, and the proposed method identified the optimal requirements. Parallel pipes must be included in the WDN solution to minimize the dissatisfaction level of the involved design. stakeholders. The objective of this work is to provide a methodology to The WDN including RO selection in complex topography design the WDN and selection of RO plants to supply has not been studied in depth. Herrera et al.16 developed a industrial plants with desalinated water. The developed model model to design water distribution networks considering RO for this methodology considers using energy recovery devices plants to supply mining plants located in high areas that are such as pumps as turbines for energy production in the isolated and arid, up to 4000 m above sea level. They used a network in places where there is a considerable difference of superstructure to design the whole system using a MINLP height that allows producing energy like in the hydropower

B DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article industry. Another novelty of the work is to considerer more 3. MATHEMATICAL FORMULATION fl than one pipeline in cases where ows are high. Four sets were used to represent the superstructure: RO plants Pumpsasturbinesareusuallyusedinurbanwater SO ={so| so is an RO plant}; nodes which can be pumping distribution networks to reduce pressure, and there are some stations, pumping stations working as turbines, or just a applications of energy production. The novelty of this work junction without a pumping stations, so N ={n|, n is a node}; relies in proposing the use of energy recovery devices in areas industrial sites or water consumers SI ={si|, si is industrial where they are usually not used, meaning areas that have a | ff site}; and the set of diameters D ={d , d is a diameter}. complex topography that are also isolated and su ering from Distances and elevations of each point were designated based water scarcity, so the only source of fresh water is desalinated on the topography data obtained using Google Earth, and water. fi Δ distances were de ned as Li,j and elevations as Zi,j. fl 2. PROBLEM STATEMENT Water ow is only in one direction, from RO plants located along the shore to the user that are industrial sites, which are Industrial sites such as mining plants are water consumers that usually mining plants. Bidirectional flow makes no sense given are located far away from the coast, in areas with a complex the high costs of transporting water from a low altitude to a topography; by complex, this refers to an area with changes in high altitude location. Only feasible connections between elevation between the RO plants and water consumers. Water nodes were considered in the WDN. demand of industrial sites can only be satisfied with desalinated Known parameters are the water requirements of industrial water from RO plants located along the coast; no other option sites, according to the needs of its facilities, water capacities of was taken into account as the industrial sites are located in RO plants, altitude of each node, and distances between each isolated areas, which means that water is a scarce resource. The node. 16,27 methodology used by Herrera et al. was used as a reference The model has equations of continuity for RO plants, nodes, in this work. and mining plants; these equations are presented as eqs 1−3 A superstructure was used to represent the WDN, which * included RO plants; pipelines; nodes with pump stations, Q so =∀∈∑ QsoSOso, n pumps as turbines, or just a junction; and industrial sites. Some nN∈ (1) nodes of the WDN are located in mountains where there is a pronounced altitude, so pipelines have to ascend and then ∑∑QQnNin,,=∀∈nj descend, so the downfall of water is used to produce energy i∈∈input joutput (2) with the passage of water through a pumping station with * pumps as turbines (PATs) like in Figure 1. The possibility of Q si =∀∈∑ QsiSInsi, nN∈ (3) fl where Qi,j is the volumetric ow pumped from i to j; constraints associated with the maximum production capacity of RO plants and desalinated water demands of the mining plants were also included in the model. 3.1. Pipe Diameter. For selecting the pipe diameter from points i to j, a disjunction expressed using the Convex Hull method28,29 was used (eq 4). Figure 1. Mountain scheme: Point A represents a node with a pump y station. Point B represents a node that is only a junction between ijd,, ∨ ∀∈(,ij ) FIJ pipes. Point C represents a PATs station and a traditional pump dD∈ DD= station. ij, d (4) ÅÄ ÑÉ TheÅ friction factorÑ is assumed to be a function of the pipe Å Ñ diameterÅ and theÑ roughness. The Reynolds number is Å Ñ consideredÇÅ big enoughÖÑ to not represent a contribution to the producing energy with water passing through the downstream 30 friction factor. A set of discrete values were considered to was evaluated and compared to water ascending through the choose diameter. The values used for the diameter were 0.7, WDN until reaching industrial sites in the traditional way of 0.8, 0.9, 1, and 1.1 m. avoiding the descend of water. 3.2. Objective Function. The objective function mini- Another consideration in the superstructure is to use more mizes the total annualized cost of desalination plants, pipes, than one pipeline until three pipelines are allowed; the choice pumping stations, and pumps as turbines required to supply between using one, two, or three pipelines depends mainly on industrial sites located in a geography with complex top- the requirements of water of each consumer’s site. There is a ography. This function includes four terms that are costs: (1) maximum velocity allowed in a pipeline, which is used as a cost of producing desalinated water, (2) costs of pumping restriction in the model. The diameters of the pipelines are stations, (3) costs of pumps as turbines stations, and (4) costs determined by the model by choosing between a set of discrete of pipelines. The fifth term is the valorization of energy diameters that are commercial sizes. produced by PATs that can be used in the grid. The objective of the model was to minimize total annual cost and find the locations and sizes of RO plants, locations TC=+++∑∑∑∑ CSO Cij, Cn CPAT and sizes of pumping stations, locations of PATs, pipe so∈∈∈∈ SO ij,, FIJ nN nN diameters, operational conditions of pumping stations, and number of pipelines in the WDN in order to provide − ∑ EPATs desalinated water to industrial sites. nN∈ (5)

C DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

The cost of producing desalinated water was extracted from operating hours in a year (h); and EC is the electricity cost the database cost raised by Wittholz et al.31 The function used ($/kW h).33 included UPC, which is the unit cost of producing desalinated To obtain the cost of pumps as turbines CPAT, the price of water. these devices was estimated using catalogs from the pumps ’ fl y company s using the net head and the ow as a reference. Civil so, c work and maintenance work of PATs was estimated using values from ref 33. ∨ CPAUPCQ=× ×* ∀∈ so so, c so so SO 3.4. Parallel Pipelines. In order to distribute the cC∈ ÅÄ ÑÉ fl Å QQQLO ≤ * ≤ UP Ñ desalinated water ow, the superstructure has the option of Å c so c Ñ (6) Å Ñ choosing up to three parallel pipelines depending on the water Å Ñ fl 34 TheÅ cost of water transport isÑ represented by a function ow. The velocity of water has a maximum value of 1.5 m/s; Å Ñ fi proposedÅ by Swamee;25 this functionÑ includes two functions: this condition is ful lled using the next equation Å Ñ Å Ñ 2 annualizedÇ capital cost of pipelinesÖ and operational and capital πD cost of pumping stations. So, the function that described the ij, Q ≤ vnj, annualized capital cost of pipelines proposed by Swamee25 is nj, 4 (11) described below The diameter of pipelines is the same for all pipelines if

Hij, m more than one is chosen. So, the next conditions must be kLDP 1 + ij,, ij fi ()Hb ful lled Cij, = ∀∈(,ij ) FIJ P (7) 3 L p Q nj, =∀∈∑ QnNnj, where kP is a proportionality constant; Hi,j is the pressure head; p=1 (12) Hb is the length parameter; Li,j is the pipe length; Di,j is the pipe diameter; and m is the exponent. H ===HHH1 2 3 The function to estimate pumping station costs is nj,, nj nj, nj, (13) The total water flow is the sum of the water flow of each ksgNb(1+ )ρ 8.76FFEDA Cρ g pipeline where p is the pipeline, and the head drop is the same Cn = + × ηPL η for every pipeline and corresponds to the total head drop. 3.5. Piecewise Linearization. The objective function is ∑ QH×∀∈ nN 3 3 ji nj, nj, zy not linear because UPC and Q are not linear, Q appears when joutput∈ j z (8) − fl j z the Darcy Weisbach equation is multiplied for the water ow, k { soQ2 is multiplied for Q. Since the objective function is not where kN is a proportionality constant; sb is the standby fraction; ρ is the mass density of water; g is the gravitational linear, the model is a MINLP model. acceleration; η is the combined efficiency of pump and prime UPC is the amortized capital cost and the operating cost of mover; F is the daily averaging factor; F is the annual obtaining desalinated water and is included in the cost of D A producing desalinated water; this cost is a compound of the averaging factor; EC is the unit electrical cost ($/kW h); and PL is the plant life years. next form The function includes the pumping head H , which is n,j CSO =× QSO UPC × cte2 (14) calculated with the Darcy−Weisbach equation, which is where Q is the desalinated water flow, and cte2 is a constant 2 SO 8fLnj, Q nj, which includes desalination plant availability and time and Hnj,0=+Δ+Hz nj , 2 5 money conversions to obtain the requirements units. π gD nj, (9) To obtain the global optimum in short time, UPC and Q3 where Δz is the elevation difference from n to j; H is the were linearized using a piecewise methodology extracted from n,j 0 Lin et al.35 Piecewise methodologies allowed us to convert a terminal head; f is the friction factor; and Ln,j is the length of the pipe from n to j. nonlinear programming problem into a linear programming 3.3. Pumps as Turbines (PATs) for Energy Production. problem or a mixed-integer convex programming problem for For pumps as turbines stations, the net head for producing obtaining an approximated global optimal solution. energy was also calculated with the Darcy−Weisbach equation extracted from ref 32. In turbines, the net head is the actual 4. CASE STUDY head used to produce energy. Net head is the gross head which An area of the Antofagasta region was used as a case study; this is the difference in height between two points minus the head area is mainly desert and isolated and features significant loss. So, in the case of pumps as turbines, the head loss would changes in elevation due to the presence of the Coastal have a minus sign in the Darcy−Weisbach equation. The Cordillera that crosses this area. The case study contained four valorization of the energy produced in a year by PATs is mining plants that can be supplied water by three potential RO calculated using the next equation: plants. Between the RO plants and the mining plants, there are N potential pumping stations, and since there are mountains in PATS USD the area, PATs were considered in some points. Connections EgQHtEPATs = ∑ ρηn nC year (10) that were unfeasible were not considered as an option. n Locations of RO plants and industrial sites were determined fl where NPATS is the number ofji PATs;yz Qn,k is the hourly ow on assumptions based on actual places where RO plants and j z through the nth PATs (m3/s);j Hn,k zis the net head (m) across mining plants could be located. Pipeline lengths were the same PATs at the samek time;{ η is the efficiency; t is determined based on real data about the water supplies of

D DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

Figure 2. Superstructure. SO are RO plants with a pumping station; n are pumping stations; P are PATs stations, and Si are industrial sites. diﬀerent mining companies. The area was chosen using Google Table 2. Distances between Each Node Earth; distances between points and elevations were Distance between nodes Distance between nodes determined using this software. Figure 2 shows the chosen Nodes (km) Nodes (km) area as a case study. RO plants are located on the coast, so their elevation is zero. n1-n4 40.7 n6-n9 22.5 There is a zone where the Coastal Cordillera goes through n1-n5 43.9 n7-n8 59.9 about 40 km of extension; this zone is where the use of PATs n2-n4 30.7 n7-n10 47.2 n2-n5 75.4 n10-n15 14.3 was considered and compared to the possibility of avoiding n2-n7 34.8 n8-n12 45.7 these mountains and to only go up until reaching the mining n2-n8 80.7 n8-n13 40.2 sites. Elevation of pumping stations varied from 0 to 2.860 km n3-n7 63.4 n11-si1 80.7 above sea level, and mining plants were located between 3.046 n3-n8 50.4 n11-si2 55.9 and 3.875 km above sea level. The elevation of each node is n4-n5 46.1 n12-si3 53.6 shown in Table 1, and the distance between each node is n4-n6 22.7 n12-si4 51.5 shown in Table 2. n5-n14 64.9 n13-si3 62.1 5. RESULTS n13-si4 49.6 n15-si2 54.3 n14-si1 55.9 n15-si3 55.0 The model was solved as a MILP problem using the Cplex n14-si2 51.2 n15-si4 76.5 solver in a GAMS environment. GAMS stands for General n14-n16 49.1 n16-si3 37.7 Algebraic Modeling System; it is a high-level modeling n14-n17 67.2 n17-si4 53.7 software for mathematical programming and optimization.

Table 1. Height Above Sea Level of Each Node GAMS is designed for modeling linear programming (LP), Height above sea level Height above sea level mixed integer linear programming (MILP), and mixed integer Nodes (km) Nodes (km) nonlinear programming (MINLP) problems.36 The Cplex SO1-n1 0 n12 2.860 solver in GAMS is designed to solve large and diﬃcult SO2-n2 0 n13 2.720 problems quickly and works to solve the majority of linear SO3-n3 0 n14 1.585 problems (LP).37 n4 2.116 n15 1.755 The global optimum was found in a short time as n5 1.364 n16 2.453 linearization allows a model to be obtained that can be solved n6 2.231 n17 2.874 quickly using Cplex, which is a desirable attribute for complex n7 2.431 Si1 3.485 models like this one. n8 1.753 Si2 3.046 5.1. Global Optimum. The requirements of desalinated n9 1.375 Si3 3.339 water of the WDN was 3.2 m3/s in total; the requirements for n10 1.765 Si4 3.875 each mining plant was 0.8 m3/s. These values were similar to n11 1.382 the requirements of the mining plants in the area.

E DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

Figure 3. Optimal solution. Solid line represents three parallel pipes, and segmented line represents two parallel pipes.

Table 3. Ranking of Solutions

Solution RO plants No. of pumping stations No. of PATs stations Energy generated by PATs (MW-h/year) Total cost (million USD/year) 1 SO2 7 1 168,593 250.71 2 SO2 7 1 168,593 251.55 3 SO2-SO3 8 1 168,593 251.78

Figure 4. Solution 2. Solid line represents three pipes, and segmented line represents two pipes.

The global optimum solution considers using only one RO year. Furthermore, considering that energy is produced from plant to supply all mining sites; in n9, there were PATs since fossil fuels, this energy production can contribute to reducing the previous node, n6, is located at a high altitude. In node n9, greenhouse gas emissions. a PATs station was followed by a traditional pump station. All A ranking of the first three optimal solutions was made to pipelines going between n2 and n11 considered using three compare them economically and operationally (Table 3). The pipelines to fulfill the water requirements of the three industrial three first solutions considered using PATs in the same area; sites (continuous line in Figure 3). Between n2 and n13, there the first two solutions considered only the SO2 RO plant to were two pipes to supply si4. When the requirements of water supply desalinated water. Instead, the third solution considered of the industrial sites were elevated, parallel pipes that came using desalinated water from SO2 and SO3. Additionally, the from one single RO plant seemed to be a suitable solution to WDN of solutions 1 and 2 used seven pumping stations, supply more than one industrial site instead of using different whereas the WDN of solution 3 used eight pumping stations. RO plants with a single or double pipeline to supply each The differences in costs between these optimal solutions were industrial site, which is the conventional way. The optimal less than one million USD/year. More details are present in solution generated 168,593 MW-h/year, which considering Table 3, and solutions 2 and 3 are illustrated in Figures 4 and 0.12 USD/kW-h represents a saving of 20.231 million USD/ 5, respectively.

F DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

Figure 5. Solution 3. Solid line represents three pipes, and segmented line represents two pipes.

Figure 6. Solution found when big changes in altitude are avoided. Solid line represents three pipes, and segmented line represents two pipes.

Figure 7. Traditional water supply system for mining plants.

G DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

Table 4. Comparison between Diﬀerent Heights of Nodes Δ Q Height n6 H n6 and n9 No. of pumping No. of PATs Total cost Energy generated by PATs (m3/s) (km) (km) RO plants stations stations (million USD/year) (MW-h/year)

3.2 2.231 0.856 SO2 7 1 250.71 168,593

3.2 2.031 0.656 SO2 7 1 254.67 127,950

3.2 1.831 0.456 SO1-SO3 10 1 257.08 7630

3.2 1.631 0.256 SO1-SO3 10 1 257.95 3931

3.2 1.431 0.056 SO1 5 0 258.52 0

By way of comparison, a network was designed without Results showed that using PATs in a region with complex considering the PATs. For this purpose, the pipes in the area topography was feasible for producing energy in the WDN by where the PATs could be installed were not considered in the passing through a region with changes in elevation like the superstructure. The optimal solution without PATs is shown in Coastal Cordillera where big changes in altitude are often Figure 6, which considered the SO1 RO plant and five avoided by mining companies. pumping stations. The cost was 258.52 million USD/year; this Linearization of two functions allowed us to find the global is 7.81 million USD/year more expensive than the option with optimum with low computational cost, which is a desired PATs. This difference was due to a higher cost in the pipes, a attribute for complex models. reduction in costs for not considering the equipment for PATs, Using parallel pipes is a feasible option to supply several and an increase in energy costs. industrial sites using water from one RO plant when the Another comparison was made to compare the optimal requirements of water are elevated, instead of having a different solution with the water supply strategy currently used by RO plant supply every industrial site with its respective mining companies, where each company had its own RO plant pipeline. and water distribution network. Since there is no energy In the future, more elements will be added to the model production, and mining companies usually avoid putting such as considering more than one water quality to supply pipelines in places with pronounced changes in elevations, multiple users and multiple sources of water to broaden the the pipelines were located where there was not a big difference water offer. Another element that can be included is using in altitude. The cost of this system (Figure 7) was 258.56 renewable energies like solar energy. million USD/year. The optimal solution was considered going through n6, ■ AUTHOR INFORMATION which was 2.231 km above sea level; the next pump station Corresponding Author coupled to a PATs station was n9, which was 1.375 km above *E-mail: [email protected]. sea level, so 0.856 km was the difference in height and 22.47 km the distance between each point. In order to analyze the ORCID Edelmira D. Galvez:́ 0000-0001-9558-443X effect of height in the optimal solution, several differences of height between n6 and n9 were used. Results are shown in Notes Table 4. The authors declare no competing financial interest. The optimal solution generated 168,593 MW-h/year, which is 20.231 million USD/year that can be utilized in the network. ■ ACKNOWLEDGMENTS If the height of the mountains was 200 m less high, the optimal The financial support from CONICYT (Fondecyt 1171341) is solution still considered energy generation with an important gratefully acknowledged. N.A. thanks CONICYT for the amount of energy generated, 127,950 MW-h/year. national Ph.D. scholarship. With a height of 1.831 km, the optimal solution considered two RO plants, SO1 and SO3, which were located at the ■ NOMENCLATURE extremes of the network, instead of choosing SO2 as the optimal solutions as in the cases of 2.231 and 2.031 km of CSO = Annualized cost of RO plants (million USD/year) height solutions. Energy production was low in comparison Ci,j = Annualized cost of pipelines (million USD/year) with the solutions obtained when elevations were higher; this is Cn = Annualized cost of pumping stations (million USD/ due to the water flow sent to n6 was 0.2 m3/s, which was very year) low in comparison to 2.4 m3/s which is the water flow sent to CPAT = Annualized cost of PATs stations (million USD/ n6 where the elevation was 2.231 km. The reason for such low year) water flow is that the net head was lower as the gross head, Di,j = Diameter of pipeline between i and j (m) which is the actual elevation of the mountains, was lower. EC = Unit electrical cost ($/(kW-h)) Head loss was high due to the distance between each point and EPAT = Valorization of energy generated by PATs in one water velocity, so the gross head had to be high to compensate year (million USD/year) for the head loss. f = Friction factor (dimensionless) FA = Annual average factor (dimensionless) F = Daily average factor (dimensionless) 6. CONCLUSIONS D g = Gravitational acceleration (m/s2) A methodology to simultaneously design a WDN to supply Hi,j = Pressure head (m) industrial sites located in regions with complex topography, Hb = Length parameter (m) fi nd the location and sizes of desalination plants, pipelines, Hn,j = Pumping head (m) pump stations, and PATs stations, was addressed. A case study H0 = Terminal head (m) was used to validate the model. Hn,k = Net head (m)

H DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

kN = Proportionality constant for pumping stations (13) Fujiwara, O.; Khang, D. B. A Two-Phase Decomposition (dimensionless) Method for Optimal Design of Looped Water Distribution Networks. Water Resour. Res. 1990, 26 (4), 539−549. kP = Proportionality constant for pipelines (dimensionless) (14) Gonzalez-Bravo,́ R.; Napoles-Rivera,́ F.; Ponce-Ortega, J. M.; Li,j = Pipe length (m) PA = Desalination plant availability (dimensionless) El-Halwagi, M. M. Multiobjective Optimization of Dual-Purpose Power Plants and Water Distribution Networks. ACS Sustainable PL = Desalination plant life (years) − fl 3 Chem. Eng. 2016, 4 (12), 6852 6866. Qi,j = Volumetric ow pumped from i to j (m /s) (15) Gonzalez-Bravo,́ R.; Fuentes-Cortes,́ L. F.; Ponce-Ortega, J. M. fl 3 Qi,n = Volumetric ow pumped from i to n (m /s) Defining Priorities in the Design of Power and Water Distribution fl 3 Qn,j = Volumetric ow pumped from n to j (m /s) Networks. Energy 2017, 137, 1026−1040. fl Qn,si = Volumetric ow pumped from n to an industrial site (16) Herrera, S.; Cisternas, L. A.; Galvez,́ E. D. Simultaneous Design (m3/s) of Desalination Plants and Distribution Water Network. Comput.- fl 3 − Qsi = Volumetric ow required in an industrial site (m /s) Aided Chem. Eng. 2015, 37 (June), 1193 1198. fl 3 Qso = Volumetric ow pumped from RO plant (m /s) (17) Tricarico, C.; Morley, M. S.; Gargano, R.; Kapelan, Z.; De fl 3 Marinis, G.; Savic,́ D.; Granata, F. Integrated Optimal Cost and Qso,n = Volumetric ow pumped from a RO plant to n(m / Pressure Management for Water Distribution Systems. Procedia Eng. s) − Q = Volumetric flow through PATs (m3/h) 2014, 70, 1659 1668. n,k (18) Nogueira Vilanova, M. R.; Perrella Balestieri, J. A. Energy and sb = Standby fraction (dimensionless) Hydraulic Efficiency in Conventional Water Supply Systems. t = Operating hours in a year (h) Renewable Sustainable Energy Rev. 2014, 30, 701−714. TC = Total annualized cost (million USD/year) (19) Zakkour, P. D.; Gaterell, M. R.; Griffin, P.; Gochin, R. J.; UPC = Unit cost of producing desalinated water (USD) Lester, J. N. Developing a Sustainable Energy Strategy for a Water vn,j = Water velocity (m/s) Utility. Part II: A Review of Potential Technologies and Approaches. − yi,j,d = Disjunction to choose diameter J. Environ. Manage. 2002, 66 (2), 115 125. ρ = Water density (kg/m3) (20) McNabola, A.; Coughlan, P.; Corcoran, L.; Power, C.; Δ ff zn,j = Elevation di erence from n to j (m) Williams, A. P.; Harris, I.; Gallagher, J.; Styles, D. Energy Recovery η = Combined efficiency of pump and prime mover in the Water Industry Using Micro-Hydropower: An Opportunity to Improve Sustainability. Water Policy 2014, 16 (1), 168−183. (21) Vilanova, M. R. N.; Balestieri, J. A. P. Hydropower Recovery in ■ REFERENCES Water Supply Systems: Models and Case Study. Energy Convers. (1) Coelho, B.; Andrade-Campos, A. Efficiency Achievement in Manage. 2014, 84, 414−426. Water Supply Systems - A Review. Renewable Sustainable Energy Rev. (22) Corcoran, L.; Mcnabola, A.; Coughlan, P. Optimization of 2014, 30,59−84. Water Distribution Networks for Combined Hydropower Energy (2) Hamiche, A. M.; Stambouli, A. B.; Flazi, S. A Review of the Recovery and Leakage Reduction. J. Water Resour. Plan. Manage. Water-Energy Nexus. Renewable Sustainable Energy Rev. 2016, 65, 2016, 142 (2), 04015045. 319−331. (23) De Marchis, M.; Fontanazza, C. M.; Freni, G.; Messineo, A.; (3) Knops, F.; Kahne, E.; de la Mata, M. G.; Fajardo, C. M. Seawater Milici, B.; Napoli, E.; Notaro, V.; Puleo, V.; Scopa, A. Energy Desalination off the Chilean Coast for Water Supply to the Mining Recovery in Water Distribution Networks. Implementation of Pumps − − as Turbine in a Dynamic Numerical Model. Procedia Eng. 2014, 70, Industry. Desalin. Water Treat. 2013, 51 (1 3), 11 18. − (4) Shenvi, S. S.; Isloor, A. M.; Ismail, A. F. A Review on RO 439 448. Membrane Technology: Developments and Challenges. Desalination (24) Carravetta, A.; Fecarotta, O.; Sinagra, M.; Tucciarelli, T. Cost- 2015, 368,10−26. Benefit Analysis for Hydropower Production in Water Distribution (5) Amit, R. K.; Ramachandran, P. Optimal Design of Water Networks by a Pump as Turbine. J. Water Resour. Plan. Manag. 2014, Distribution Networks: A Review; 2009. 140 (6), 04014002. (6) Khor, C. S.; Chachuat, B.; Shah, N. Optimization of Water (25) Swamee, P. K. Optimal Design of Pumping Mains. J. Transp. Eng. 2001, 127 (3), 264−267. Network Synthesis for Single-Site and Continuous Processes: (26) Gupta, R.; Baby, A.; Arya, P. V.; Ormsbee, L. Upgrading Milestones, Challenges, and Future Directions. Ind. Eng. Chem. Res. Reliability of Water Distribution Networks Recognizing Valve 2014, 53 (25), 10257−10275. Locations. Procedia Eng. 2014, 89, 370−377. (7) El-Halwagi, M. M. Synthesis of Reverse-osmosis Networks for ́ − (27) Herrera-Leon, S.; Lucay, F.; Kraslawski, A.; Cisternas, L. A.; Waste Reduction. AIChE J. 1992, 38 (8), 1185 1198. Galvez,́ E. D. Optimization Approach to Designing Water Supply (8) Liu, S.; Konstantopoulou, F.; Gikas, P.; Papageorgiou, L. G. A Systems in Non-Coastal Areas Suffering from Water Scarcity. Water Mixed Integer Optimisation Approach for Integrated Water Resour. Manage. 2018, 32 (7), 2457. Resources Management. Comput. Chem. Eng. 2011, 35 (5), 858−875. ́ (28) Iqbal, K. Fundamental Engineering Optimization Methods; (9) Ahmetovic, E.; Grossmann, I. E. Global Superstructure Bookboon.com, 2013; Vol. 53. Optimization for the Design of Integrated Process Water Networks. (29) Hendrix, E. M. T.; G.-Toth,́ B. Introduction to Nonlinear and − AIChE J. 2011, 57 (2), 434 457. Global Optimization; Springer, 2010; Vol. 37. (10) Atilhan, S.; Mahfouz, A. B.; Batchelor, B.; Linke, P.; Abdel- (30) Swamee, P. K.; Jain, A. K. Explicit Equations for Pipe Flow ́ ́ ́ Wahab, A.; Napoles-Rivera, F.; Jimenez-Gutierrez, A.; El-Halwagi, M. Problems. J. Hydraul. Div., Am. Soc. Civ. Eng. 1976, 102, 657−664. M. A Systems-Integration Approach to the Optimization of (31) Wittholz, M. K.; O’Neill, B. K.; Colby, C. B.; Lewis, D. Macroscopic Water Desalination and Distribution Networks: A Estimating the Cost of Desalination Plants Using a Cost Database. General Framework Applied to Qatar’s Water Resources. Clean Desalination 2008, 229 (1−3), 10−20. Technol. Environ. Policy 2012, 14 (2), 161−171. (32) Swamee, P. K.; Sharma, A. K. Design of Water Supply Pipe (11) Lira-Barragan,́ L. F.; Ponce-Ortega, J. M.; Guillen-Gosá lbez,́ G.; Networks; Wiley, 2008. El-Halwagi, M. M. Optimal Water Management under Uncertainty for (33) Fontana, N.; Giugni, M.; Portolano, D. Losses Reduction and Shale Gas Production. Ind. Eng. Chem. Res. 2016, 55 (5), 1322−1335. Energy Production in Water-Distribution Networks. J. Water Resour. (12) Liang, Y.; Pahija, E.; Hui, C. Convex Model for Global Plan. Manag. 2012, 138 (3), 237−244. Optimization of Water Distribution System. Comput.-Aided Chem. (34) Water Pipeline Design Guidelines; Saskatchewan Environment, Eng. 2016, 38, 493−498. April 2004.

I DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX Industrial & Engineering Chemistry Research Article

(35) Lin, M.-H.; Carlsson, J. G.; Ge, D.; Shi, J.; Tsai, J.-F. A Review of Piecewise Linearization Methods. Math. Probl. Eng. 2013, 2013,1− 8. (36) Rosenthal, R. E. GAMS  A User’sGuide;GAMS Development Corporation, March 2016. (37) IBM ILOG CPLEX Optimization Studio CPLEX User’s Manual; IBM Corp., 2016.

J DOI: 10.1021/acs.iecr.7b05247 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

PUBLICATION II

Araya, N., Kraslawski, A., Cisternas, L.A.

Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings.

Journal of Cleaner Production

Vol 263, 2020

https://doi.org/10.1016/j.jclepro.2020.121555 Journal of Cleaner Production 263 (2020) 121555

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Towards mine tailings valorization: Recovery of critical materials from Chilean mine tailings

* Natalia Araya a, b, , Andrzej Kraslawski b, Luis A. Cisternas a a Departamento de Ingeniería Química y Procesos de Minerales, Universidad de Antofagasta, Antofagasta, 1240000, Chile b School of Engineering Science, Lappeenranta-Lahti University of Technology (LUT University), Lappeenranta, FI-53851, Finland article info abstract

Article history: The mining industry produces large volumes of mine tailings e a mix of crushed rocks and process ef- Received 16 June 2019 fluents from the processing of mineral ores. Mine tailings are a major environmental issue due to im- Received in revised form plications related to their handling and storage. Depending on the mined ore and the process used, it 6 March 2020 may be possible to recover valuable elements from mine tailings, among them critical raw materials Accepted 5 April 2020 (CRMs) like rare earths, vanadium, and antimony. Available online 9 April 2020 The aim of this study was to investigate the techno-economic feasibility of producing critical raw Handling editor:Yutao Wang materials from mine tailings. Data from 477 Chilean tailings facilities were analyzed and used in the techno-economic assessment of the valorization of mine tailings in the form of CRMs recovery. A review Keywords: of applicable technologies was performed to identify suitable technologies for mine tailings processing. Mine tailings To assess the economic feasibility of CRMs production, net present value (NPV) was calculated using the Critical raw materials discounted cash flow (DCF) method. Sensitivity analysis and design of experiments were performed to Techno-economic assessment analyze the influence of independent variables on NPV. Two options were assessed, rare earth oxides Discounted cash flow (REOs) production and vanadium pentoxide (V2O5) production. The results show that it is possible to Sensitivity analysis produce V2O5 with an NPV of 76 million US$. In the case of REOs, NPV is positive but rather low, which indicates that the investment is risky. Sensitivity analysis and the ANOVA run using the design of experiments indicated that the NPV of REOs is highly sensitive to the price of REOs and to the discount rate. © 2020 Elsevier Ltd. All rights reserved.

1. Introduction serious contamination of soils and groundwater with nearby communities particularly badly affected by the results of eolian and Mine tailings are waste from the processing of mineral ores. water erosion of tailing disposal sites (Mendez and Maier, 2008). They are a mixture of ground rocks and process effluents generated Another cause of environmental pollution from mine tailings is acid during processing of the ores, and their composition depends on mine drainage (AMD) (Larsson et al., 2018; Moodley et al., 2017). the nature of the mined rock and the recovery process used. In AMD is formed from the exposure of sulfide ores and minerals to copper mining, tailings can account for 95e99% of crushed and water and oxygen, once the ore is exposed, sulfate and heavy ground ores (Edraki et al., 2014). Worldwide, mine tailings are metals are released into the water (Moodley et al., 2017). AMD is produced at a rate of anywhere from five to fourteen billion tons considered one of the most significant forms of water pollution and per year (Adiansyah et al., 2015; Edraki et al., 2014; Schoenberger, the USA Environmental Protection Agency (US-EPA) considers it to 2016). be the second only to global warming and ozone depletion in terms In view of the volumes of mine waste produced and the nature of ecological risk (Moodley et al., 2017). of the chemicals involved, the storage and handling of mine tailings Tailing storage facilities (TSF), also called tailing deposits, are the is a significant environmental problem. Mine tailings are a source of source of most mining-related disasters (Schoenberger, 2016). Ap- proaches to the handling and storage of mine tailings include riverine disposal, wetland retention, backfilling, dry stacking and storage behind damned impoundments (Kossoff et al., 2014). Mine * Corresponding author. School of Engineering Science, LUT University, FI-53851, tailings dam failures can have catastrophic consequences. 237 cases Lappeenranta, Finland. fi E-mail addresses: natalia.araya.gomez@lut.fi (N. Araya), andrzej.kraslawski@lut. of signi cant tailings accidents were reported for the period 1971 to fi (A. Kraslawski), [email protected] (L.A. Cisternas). 2009 (Adiansyah et al., 2015). More recently, in January 2019, an https://doi.org/10.1016/j.jclepro.2020.121555 0959-6526/© 2020 Elsevier Ltd. All rights reserved. 2 N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 accident at the Corrego do Feijao~ mine in Brumadinho in the since 2011, the last update was in 2017, and it currently contains metropolitan region of Belo Horizonte in southeastern Brazil killed twenty-seven CRMs including 3 element groups: light rare earth at least 65 people with about 280 people were missing (De Sa, elements (LREEs), heavy rare earth elements (HREES) and platinum 2019). group elements. To achieve a circular economy model, the valorization of mine According to the International Union for Pure and Applied tailings is crucial for the mining industry, which needs to improve Chemistry (IUPAC), rare earth elements (REEs) are a group of 17 its processes to minimize its environmental impact and close the elements that includes lanthanides, composed of 15 elements, and loops (Kinnunen and Kaksonen, 2019). Different approaches to yttrium and scandium, which are included in this group due to the tailings valorization can be taken, such as reprocessing to extract similarity in chemical characteristics. REEs can be found in over 250 metals and minerals, tailings as backfill material, tailings as con- different minerals (Jordens et al., 2013; Sadri et al., 2017). REEs have struction material, energy recovery and carbon dioxide sequestra- an important role in the transition to green technologies because of tion (Lottermoser, 2011). their use in crucial components such as permanent magnets and Challenges that the mining industry needs to face to achieve the rechargeable batteries, and their use as catalysts (Binnemans et al., valorization of tailings aligned with circular economy principles 2013a). China is responsible for almost 80% of the global supply of include improving the rather limited knowledge about mineralogy, REEs, such monopoly has raised concerns about a possible shortage impurities concentration, and the quantity of tailings; developing of supply, (Hornby and Sanderson, 2019; Vekasi and Hunnewell, new business models that take account of price development, 2019). lower disposal costs, and market demand; providing institutional Other elements on the list of CRMs are platinum group elements impulse indispensable to encourage the transformation from a (PGEs), which include ruthenium (Ru), rhodium (Rh), palladium linear to a circular economy; technology development to make (Pd), osmium (Os), iridium (Ir) and platinum (Pt). These metals are processes economically feasible since most mine tailings have low very rare in the Earth’s continental crust, ranging from 0.022 ppb grades of different elements mixed with residues of previous pro- for iridium to 0.52 to Pd (Mudd et al., 2018). cesses (Kinnunen and Kaksonen, 2019; Lottermoser, 2011). Nowadays, due to the increasing demand for CRMs, new sources Due to the geological heterogeneity of the rocks mined and the are being sought, and secondary sources such as metal scrap and continuous flow processes used in mineral processing, tailings industrial waste are attracting more attention. The use of the deposits contain large quantities of valuable elements whose re- hitherto unexploited secondary sources can reduce demand for covery could bring potential economic benefits. A number of virgin materials and, in consequence, contribute to a decrease in studies have investigated the recovery of valuable elements from mining production. One of the core principles of the circular mine tailings (Ahmadi et al., 2015; Alcalde et al., 2018; Andersson economy is the reduction and minimization of resource use, and et al., 2018; Ceniceros-Gomez et al., 2018; Falagan et al., 2017; ways to achieve that goal include recycling and reuse of wastes Figueiredo et al., 2018; Khalil et al., 2019; Khorasanipour, 2015; (Kirchherr et al., 2017). Mine tailings from mineral processing of a Mohamed et al., 2017; Sracek et al., 2010). certain branch of the metal industry could be used as a source in a As shown by recent studies (Ceniceros-Gomez et al., 2018; process designed to obtain one or more critical raw materials, a Markovaara-Koivisto et al., 2018; Moran-Palacios et al., 2019; Tunsu simplified flowsheet of this idea is shown in Fig. 1. et al., 2019), among elements contained in mine tailings, there are Chile has a long history of mining and large-scale mining started many critical raw materials (CRMs). Raw materials have significant in the first decade of the twentieth century. In 2016, Chilean mining economic importance and are utilized in the manufacture of a wide exports were valued at 30,379 million USD according to the Na- range of goods. In particular, critical raw materials can be applied in tional Service of Geology and Mining (SERNAGEOMIN), 90% of areas such as alternative energy production and communications which came from copper mining (SERNAGEOMIN, 2017). Chile is devices, and they play a significant role in the development of the world’s leading producer of copper. Currently, a decrease in the globally competitive and eco-friendly innovations. Securing access grade of mined copper ores is being observed, which increases the to a stable supply of many raw materials has become a major amount of processed ore and, consequently, leads to greater tailings challenge for national and regional economies with a limited pro- deposits for the same level of copper production. Currently, Chile duction, which relies on imports of numerous minerals and metals produces 1,400,000 tons of mine tailings daily and there are 696 (European Commission, 2017a). tailings storage facilities (TSF) (SERNAGEOMIN, 2018). Many studies have examined the criticality of raw materials. The objective of this study is to conduct a technical and eco- This study utilizes the list compiled by the European Commission nomic assessment of the valorization of mine tailings of Chile as a (EC), where raw materials are considered critical when they are source of CRMs. Therefore, the research questions addressed in this both of high economic importance for the European Union (EU) and paper are: vulnerable to supply disruptions (European Commission, 2017b). What critical materials can be recovered from mine tailings? The term “vulnerable to supply disruption” means that their supply What are the challenges in the production of critical materials is associated with a high risk of not meeting the demand of the EU using mine tailings as a source? industry. High economic importance means that the raw material is In recent years, the use of secondary sources for obtaining raw of fundamental importance to industry sectors that create added materials has gained growing importance. This research supple- value and jobs, which may be lost in the case of inadequate supply ments these works with a techno-economic feasibility study for and if adequate substitutes cannot be found (Blengini et al., 2017). producing critical raw materials from mine tailings. The most critical metals are those for which supply constraints The data used in the study refer to mine tailings samples of 477 result from the fact that they are largely or entirely mined as by- Chilean copper mining industrial deposits. These data have not products, generate environmental impacts during production, been previously used to assess the economic potential of the re- have no effective substitutes, and are mined in areas prone to covery of critical materials. geopolitical conflict (Graedel et al., 2015). In 2011, the European Commission (EC) published a list of 14 raw 2. Methodology materials that are critical for emerging technologies of European industries, so-called critical raw materials (CRMs) (European The first step to evaluate the recovery of CRMs from mine tail- Commission, 2017a, 2014, 2011). The list has been updated twice ings is the calculation of the amount of each CRM present in N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 3

Fig. 1. Simplified mining processes flowsheet featuring conventional processes to obtain metal and the re-processing of tailings to obtain CRMs. tailings. The feasibility of recovery is next assessed for critical tailing deposit in the country (Ghorbani and Kuan, 2017). Previ- materials found in larger quantities. ously, prior to the adoption of appropriate regulations, tailings were In the technological assessment, technologies for processing abandoned in deposits and no efforts were made to ensure the mine tailings are first examined. If no technologies are available, safety of the nearby communities but nowadays the handling and technologies for processing ore, as an analogous process, are storage of tailings are strictly regulated. In 2011, the Law 22.551 was considered taking into account differences between the processing promulgated. It regulates the closing of mining facilities and of ore and processing of waste. specifies that tailings must be physically and chemically stabilized In the economic assessment, the discounted cash flow (DCF) (Ministerio de Minería, 2011; SERNAGEOMIN, 2011). method is used to assess the feasibility of the options for the re- In Chile, there are 696 mine tailings deposits registered in a covery CRMs from mine tailings. This method has been widely used national registry, compiled between 2016 and 2018. The registry is for valuation projects (De Reyck et al., 2008; Kodukula and expected to be updated as new mine tailings facilities are opened Papudesu, 2006; Zizlavský, 2014). DCF is a commonly adopted and old abandoned tailing deposits are discovered. Antofagasta economic valuation technique and consists of discounting expected Region hosts larger mine tailings deposits (SERNAGEOMIN, 2018) cash flow of a future project at a given discount rate and then because of the size of the mining sector in this region, which ac- summing all the cash flows of a determined period of time (Ibanez-~ counts for 47% of the contribution to Chilean mining activity. The Fores et al., 2014; Zizlavský, 2014). most serious problems associated with tailings and handling and Sensitivity analysis is performed to assess the impact of various storage of tailings are related to the seismic nature of the country, parameters on the NPV of CRMs recovery from mining tailings. and risks associated with tailings dam failure include fatalities, Sensitivity analysis is a tool used to analyze how different values of serious water contamination, and destruction of the land. a set of independent variables affect a dependent variable. The sale price of critical materials, operating costs, capital costs, and discount rate are the main inputs in the DCF method, then these 3.1. Characterization of mine tailings variables are studied in the sensitivity analysis. These variables and interactions among them were also tested using a design of ex- The chemical composition of tailings in 477 mine tailings de- periments with response surface methodology. posits is available on the website of the National Service of Geology and Mining of Chile (SERNAGEOMIN) (SERNAGEOMIN, 2018). This database contains values for concentrations of 56 elements, 3. Mine tailings assessment including 22 CRMs featuring on the latest EC list. The CRMs analyzed in the SERNAGEOMIN database are vanadium, cobalt, Mining is one of the main economic activities in Chile due to the yttrium, niobium, scandium, hafnium, tantalum, antimony, bis- country’s favorable geochemical and mineralogical characteristics. muth, tungsten, lanthanum, cerium, praseodymium, neodymium, Chile is the world’s leading producer of copper, producing 5,552.6 samarium, europium, gadolinium, terbium, dysprosium, holmium, thousand tons of copper in 2016 (SERNAGEOMIN, 2017), the world’s erbium, thulium, ytterbium and lutetium (SERNAGEOMIN, 2018). second supplier of molybdenum, producing 62,746.1 tons in 2017, Chemical composition in each mine tailings deposit is different and the second producer of lithium, producing 77,284 tons of and it depends on the type of mineral rocks mined and the pro- lithium carbonate in 2017 (SERNAGEOMIN, 2017). For some regions cesses used in the plant. In the geochemical characterization of in Chile, mining is the main economic activity; most mining activity Chilean tailings, it can be noticed that most tailings deposits have a is found in the Atacama Desert in northern Chile. high percentage of silicon oxide or ferric oxide due to the type of The Atacama Desert is the driest non-polar desert on the earth, minerals processed (SERNAGEOMIN, 2018). and its copper ore deposits are world-class porphyry copper de- Data in the SERNAGEOMIN database are classified by the current posits (Oyarzún et al., 2016; Tapia et al., 2018). Porphyry deposits status of the tailings deposits: active, inactive, and abandoned. In the are the principal sources of copper and molybdenum methodology used in this study, only inactive and abandoned tailings (Khorasanipour and Jafari, 2017). Porphyry deposits consist of were analyzed, because their volume and chemical composition do distributed and stockwork sulfide mineralization located in various not change over time. In the case of active tailings, although their host rocks that have been altered by hydrothermal solutions into volume is greatest, their chemical composition may change over the roughly concentric zonal patterns (Dold and Fontbote, 2001). course of years, which is why they have not been considered in this Chilean mining processing plants produce large quantities of study. Mine tailings of the Antofagasta Region are examined because waste every year. Tailings dams are the most common type of the tailing volume storage is greater in this region than in other 4 N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 regions. The TSFs analyzed cover 16 inactive deposits. The location of are assumed to be also applicable to the processing of mine tailings. mine tailings of the Antofagasta Region can be seen in Fig. 2. Based on the content of the mine tailings analyzed, two feasibility CRMs found in larger quantities are given in Table 1. The sum of studies are conducted; the first for producing rare earth oxides and REEs was also calculated, to produce REE concentrate or mis- the second for vanadium recovery, using mine tailings as a source. chmetal, which is an alloy of REEs. The sum of REEs does not The extraction process for REEs, in a general form, includes three consider scandium because it is separated in a different process. steps: mining and comminution; ore beneficiation processes consisting of flotation, gravity and magnetic techniques to generate 3.2. Technology assessment REE concentrate; and hydrometallurgical methods to extract REE compounds (Sadri et al., 2017). Hydrometallurgical methods A literature review was conducted to investigate the available include cracking of REE concentrate; leaching, neutralization and fi technologies for the recovery of critical raw materials from mine precipitation processes; and separation and puri cation techniques tailings. If no technologies are available for tailings processing, then such as solvent extraction. Solvent extraction allows recovering those used for processing of primary ores are considered as a REEs with a high degree of purity, moreover, a variety of solvent reference. It is important to notice that mine tailings are already in extraction reagents is available. For secondary waste, selective the form of slurry or paste, depending on the percentage of water extraction of REEs is required from solutions with a high content of present, so there are no mining costs, which represent approxi- other species (Tunsu et al., 2019). mately 43% of operating cost in a mine (Curry et al., 2014). A life cycle inventory and impact assessment of the production Existing technologies for CRMs production are briefly described of RE oxides from primary bastnasite and monazite has been pre- in Table 2. Most of these technologies are for primary ores. Some sented for the Bayan Obo mine in Inner Mongolia, China, in (Koltun applications for secondary sources such as industrial waste and and Tharumarajah, 2014). The study found out the mining and fi mine tailings exist (Abisheva et al., 2017; Binnemans et al., 2015; bene ciation stage accounts for 6.98% of energy consumption and Figueiredo et al., 2018; Innocenzi et al., 2014; Jorjani and Shahbazi, 6.51% of water consumption. When processing mine tailings, there 2016; Peelman et al., 2016), but they should be treated as emerging is no mining stage, so the values were adapted. Adapted values of technologies. Significant further development of these new tech- energy and water consumption to obtain RE oxides from waste nologies is required before they are suitable for industrial-scale material are included in the supplementary material. usage (Kinnunen and Kaksonen, 2019). Primary ores of REEs are usually treated with alkaline pressure In spite of the low concentration of REEs in comparison to end- leaching or sulfuric acid roasting. However, mine tailings are a low- of-life consumer goods, mine tailings are a potential source of REEs grade source of REEs, so these technologies may not be economically because of the large volumes of mine tailings, which mean that the feasible. Chloride-based hydrometallurgical processes may be a total amount of recoverable REEs could be high (Binnemans et al., potential alternative to traditional capital intensive hydrometallur- 2015). Several processes have been proposed for the recovery of gical processes based on high temperature and pressure (Onyedika REEs from mine tailings. Peelman et al. (2018) have proposed a et al., 2012) and they could be a suitable option for REE recovery method for the recovery of REEs from mine tailings from apatite from tailings at economically viable capital and operating cost. mineral with an REE content of 1200e1500 ppm using acidic In the case of vanadium, it is mainly produced as a co-product leaching followed by cryogenic crystallization and solvent extrac- from the vanadium slag before the steel converter. The main va- tion. They achieved a 70e100% recovery of REE. nadium products are vanadium pentoxide (V2O5) and ferrovana- There are no processing plants using copper mine tailings as a dium (FeV) (European Commission, 2017b). Other sources of source of CRMs. Therefore, technologies used for primary sources vanadium are stone coal, steel scrap, and fossil fuels. The mine tailings analyzed in this study have a CRMs content that varies between 80 and 214,000 g per ton of tailing. In Chile, there are currently no projects providing for the use of mine tailings as a source of CRMs, nor approved initiatives for the production of CRMs from primary ores.

3.3. Economic assessment

The economic assessment is done in two main steps. The ﬁrst step focuses on the economic potential of CRMs found in inactive mine tailings as an in-situ value, considering the monetary value of the CRMs to assess the feasibility of CRMs production. The second stage concentrates on the analysis of the feasibility of CRMs production using mine tailings as a source. Prices of critical materials may differ from one source to another. In addition, the prices of some critical materials are not publicly available as they are traded privately. To calculate the economic potential of inactive mine tailings deposits, the following prices were used, see Table 3. The economic potential of CRMs recovery was calculated as the fraction of each CRM in the tailings multiplied by the mass of each TSF for the 16 TSFs studied. The economic potential is a reference value for the total REE value of the mine tailings. The economic potential of these TSFs is shown as supplementary material. To assess the feasibility of CRMs recovery, the DFC method was Fig. 2. Tailings storage facilities in Antofagasta Region, blue represents inactive or abandoned deposits and red is for active deposits. (For interpretation of the references used to calculate the NPV and IRR for REOs production and V2O5 to colour in this ﬁgure legend, the reader is referred to the Web version of this article.) production using mine tailings. The NPV is the difference between the N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 5

Table 1 Total tonnage and uses of CRMs present in inactive tailings deposits of the Antofagasta Region (16 deposits).

CRMs Tons Uses

Vanadium (V) 46,110 Most of the vanadium produced is used in ferrovanadium or as a steel additive. Another use is as vanadium pentoxide. Cerium (Ce) 22,886 Cerium is used as a catalyst converter for carbon monoxide emissions, as an additive in glass for reducing UV transmission, and in carbon-arc lighting. Cobalt (Co) 16,940 The main uses of cobalt are in battery chemicals for NieCd, Ni-metal hydride and Li-ion battery types, superalloys, hard materials, catalysts, and magnets. Yttrium (Y) 16,039 Yttrium is used for energy-efficient fluorescent lamps, in the treatment of various cancers, in aerospace surface and barriers, as a superconductor, in aluminum and magnesium alloys, and in-camera lenses. Neodymium 14,880 Neodymium is used to create high-strength magnets for computers, cell phones, medical equipment, electric cars, wind turbines, and audio (Nd) systems. It is also used in the glass and ceramic industries. Lanthanum (La) 10,253 Lanthanum is used in nickel metal hydride rechargeable batteries for hybrid automobiles, in high-quality camera and telescope lenses, and in petroleum cracking catalysts in oil refineries. Scandium (Sc) 9,359 Scandium is used to increase strength and corrosion resistance in aluminum alloys, in high-intensity discharge lamps, and in fuel cells to increase efficiency at lower temperatures. Niobium (Nb) 4,823 Niobium is used in high strength low alloy (HSLA) steels as ferroniobium and in superconducting magnets. Antimony (Sb) 3,751 Principal uses for antimony are in alloys with lead and tin, and in lead-acid batteries. Samarium (Sm) 3,456 The main use of samarium is in cobalt-samarium alloy magnets for small motors, quartz watches, and camera shutters. Samarium is also used in lasers. Gadolinium 3,357 Gadolinium is mainly used for NdFeB permanent magnets, lightning applications and in metallurgy. (Gd) Praseodymium 3,245 Praseodymium is used in NdFeB magnets, ceramics, batteries, catalysts, glass polishing and fiber amplifiers. (Pr) Dysprosium 2,705 Dysprosium is used mainly and almost inclusively in NdFeB magnets. (Dy) REEs (total) 82,254

Table 2 Available and emerging technologies for CRMs processing.

CRMs Production process

Rare earth - Acidic leaching-cryogenic crystallization-solvent extraction from mine tailings with apatite and monazite. (Peelman et al., 2016). elements - Bioleaching for REEs extraction from low-grade sources. (Peelman et al., 2014). - Solvent extraction to recover REEs from mine tailings of gold and tellurium mining (Tunsu et al., 2019). -Use of solvent impregnated resins (SIR) to recover REEs from low concentration solutions (Onishi et al., 2010; Sun et al., 2009; Yoon et al., 2016) Antimony - Crushing and pyrometallurgical methods for primary ores (Anderson, 2012). -Crushing and hydrometallurgical methods like leaching and electrodeposition (Anderson, 2012) Cobalt - Bioleaching of sulfidic tailings of iron mines. (Ahmadi et al., 2015). - Mineral beneficiation, comminution, flotation, smelting, leaching or refining for sulfide ores (European Commission, 2017b). -Calcination, pyrometallurgical process, hydrometallurgical methods for lanthanides ores (European Commission, 2017b) Niobium -Gravity separation, froth flotation, magnetic and electrostatic separation, and acid leaching depending on the ore (European Commission, 2017b) Vanadium - Extraction of vanadium as a co-product to iron from vanadium slag includes bearing, roasting, acid leaching solvent extraction, ion exchange, and precipitation (Xiang et al., 2018). - Desliming-flotation from low-grade stone coal (European Commission, 2017b).

-Preform reduction process (PRP) based on a metallothermic reduction of vanadium pentoxide (V2O5). (Miyauchi and Okabe, 2010).

Table 3 CRMs prices in July 2018.

Critical material Price ($US/kg)a Critical material Price (US$/kg)a

Antimony 8.51 Neodymium metal 99.5% 68.0 Cerium metal 99.5% 7.00 Neodymium oxide 99.5% 66.7 Cerium oxide 99.5% 5.59 Praseodymium metal 99% 125.00 Cobalt 87.5 Praseodymium oxide 99.5% 81.6 Dysprosium metal 99% 268.57 Samarium metal 99.9% 15 Dysprosium oxide 99.5% 226.80 Scandium metal 99.9% 3,458 Gadolinium metal 99.9% 44.00 Scandium oxide 99.95% 1,079 Gadolinium oxide 99.5% 20.94 Vanadium (as V2O5 80%) 40.00 Lanthanum metal 99% 7.00 Yttrium metal 99.9% 36.5 Lanthanum oxide 99.5% 7.80 Yttrium oxide 99.99% 4.60

a Sources: (Mineralprices.com, 2018), (Thenorthernminer.com, 2018), (LME, 2018). present value of cash inflows and the present value of cash outflows in represent the investment made for the project, which includes a particular period of time. IRR is the discount rate at which the NPV of costs of the development phase which, among other costs, com- future cash flows is equal to the initial investment. NPV and IRR are prises the purchase of the equipment, building a manufacturing metrics used in capital budgeting and decision-making. The calcula- plant and the cost of product launch. The investment represents the tion does not include external factors such as inflation. To obtain the first cash flow in the DFC method. NPV and IRR for the options assessed, capital costs and operating costs Operating costs, operating expenses or OPEX, are expenses of projects with similar characteristics were used. incurred during the lifetime of the project. In the case of a mining Capital costs, also referred to as capital expenses or CAPEX, project, these would include the cost of labor, water, and energy, 6 N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 maintenance, spare parts, and indirect costs (Bhojwani et al., 2019). vanadium project (Lee, 2018). This project has been designed as an The first option assessed is the production, using mine tailings open pit heap leaching operation to obtain vanadium pentoxide as a source, of the following rare earth oxides (REOs): cerium, (V2O5). The Gibellini project is designed for processing of low-grade lanthanum, neodymium, yttrium, samarium, gadolinium, praseo- minerals, so it is suitable for mine tailings, but in this study, pro- dymium and dysprosium. Scandium is also considered as REE but it duction is reduced because the grade in mine tailings is lower in has different properties and a different production process, which mine tailings. The values used for the calculation of NPV and IRR are is why it was not assessed together with the above mentioned REEs. given in Table 5. The values of NPV and IRR for vanadium produc- The second option assessed is vanadium as the production of tion from Chilean mine tailings are shown in the supplementary vanadium pentoxide (V2O5). It is due to the fact that vanadium is material. the main CRM found TSFs in the Antofagasta Region (see Table 1). The NPV is 76 million US$ and the IRR is 21%, these values indicate that the project is profitable as the NPV is positive and the IRR is higher than the discount rate. Cash inflows and outflows are 3.3.1. Feasibility of producing rare earth elements using mine shown as supplementary material. tailings as a source For REOs production, we have considered only REEs found in larger quantities. Due to the lack of data about similar projects that 3.4. Sensitivity analysis use mine tailings or industrial waste as source material, we used data from a Canadian project that produces rare earth oxides In this study, a sensitivity analysis was performed on four pa- (Hudson Resources Inc, 2013) from primary sources to produce of rameters: capital cost, operating cost, critical materials price, and neodymium, praseodymium, lanthanum, and cerium. Data used for the effect of the discount rate on NPV for the examined options. The NVP calculation are shown in Table 4. objective of the sensitivity analysis is to understand the uncertainty The price used to calculate NPV corresponds to the weighted in the NPV for the examined parameters. These parameters were average for REOs; cerium, lanthanum, samarium, gadolinium, chosen because they are the key components in the DCF method. praseodymium, dysprosium, and yttrium oxide, which is 37 USD/kg Sensitivity analysis determines how different values of one or of REOs produced, 40% was discounted to reflect the difference more independent variables affect a dependent variable under a between REO concentrate and separated individual rare earth oxide given set of assumptions. Sensitivity analysis is the last stage of the prices, so the price used for NPV calculations is 22 USD/kg, as in the process of assessing and selecting a technological alternative report it was used as a reference price. The grade of REEs corre- (Ibanez-For~ es et al., 2014). Sensitivity analysis studies how several sponds to the average REEs grade in all the deposits analyzed. In the sources of uncertainty contribute to the entire uncertainty of a mine tailings covered by the analysis, the average grade is lower mathematical model. than in most primary ore processing projects, so the production In the DCF method, the discount rate is the rate used to convert was reduced accordingly. the future value of a project cash flows to today’s value. The dis- It is important to note that operating costs and capital costs are count rate is adjusted to the risk associated with a project. There- referential values. In the case of mine tailings, costs related to fore, the higher the risk, the higher the discount rate (Kodukula and extracting mineral ores should not be considered since tailings are Papudesu, 2006). Risk is associated with the uncertainty of a materials that have already been mined and processed. project. In business, risks may have a positive or negative effect. The The NPV is 672,987 USD which means that the projected earn- discount rate was varied to acknowledge that mining projects deal ings generated for this proposed REOs production exceed the with uncertainties that can be included in the model by choosing a anticipated costs and the overall value for the project is positive. higher discount rate. However, even though the NPV is positive, its value is too low to Mining commodity prices always show greater volatility than invest in a project of such a magnitude. The IRR is 10.03% which is those of any other primary products (Foo et al., 2018). Prices of almost the same as the discount rate chosen for the project, this critical materials may experience price spikes due to their insta- confirms that the project is not highly profitable. Cash inflows and bility caused by the risk of supply disruption. Critical materials have outflows are included as supplementary material. inelasticity element in their prices, this means that the demand for these materials is not highly affected by the price (Binnemans et al., 3.3.2. Feasibility of producing vanadium using mine tailings as a 2013b; Leader et al., 2019). Critical materials are needed in tech- source nologies, such as clean energy technologies, in which there are not Vanadium is the main CRM found in mine tailings in the Anto- substitutes for the critical materials needed (Leader et al., 2019). fagasta Region. There are 46,110 tons of vanadium in inactive TSFs, The price of each critical material assessed was considered as an but active tailings in this area have the potential for ca. 900, 000 important parameter that contributes to the overall uncertainty of tons of vanadium. the project. Capital and operating costs for vanadium production are taken Since capital costs and operating costs used in this study are from a preliminary economic assessment study for the Gibellini referential values, and they are further used as inputs in the DCF

Table 4 Table 5 Data for REOs project. Data for vanadium project.

Data Value Unit Data Value Unit

Capital cost 342,514,448 US$ Capital cost including 25% contingency 116,760,000 US$ Life of mine 20 years Life of mine 14 years

Operating cost 13,080 US$/ton REOs Operating cost 14,767 US$/ton V2O5 REOs Price 22,000 US$/ton REOs Vanadium pentoxide price 40,000 US$/ton V2O5 Production capacity 4,000 tons REOs/year Production capacity 1,000 tons V2O5/year Annual increase (OPEX) 1.5 % Annual increase (OPEX) 1.5 % Annual increase (PRICE) 1.5 % Annual increase (PRICE) 1.5 % Discount rate 10 % Discount rate 10 % N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 7 method, it was necessary to address the variability of the real values shows that if the discount rate increases by 0.01 from the value of of these parameters vis-a-vis the values used here. 0.1 used to 0.11, the NPV decreases by 10 million US$ approximately. Capital cost, operating costs, and prices varied between 30 and When the discount rate is higher than 0.21, NPV becomes negative. 30% of the original value. The discount rate varied between 0.05 and Results show that under certain prices, operating costs and 0.3. capital costs, it is possible to invest in producing CRMs using a The results of the sensitivity analysis for the REOs price are secondary source such as mine tailings. shown in Fig. 3. It can be seen that for every 5% increase in the price The parameters analyzed in the sensitivity study may change of the REOs, the NPV increases by 38 million US$. NPV is highly simultaneously. Therefore, their interactions were analyzed using sensitive to changes in REO prices. NPV becomes negative when the design-of-experiments together with response surface methodol- price of REOs is below 22 US$/kg, making the project financially ogy. In the analysis of the NPV of both projects, REOs production unviable. and V2O5 production, four factors and three levels were considered. The NPV is less sensitive to changes in operating costs than The factors are: the price, capital costs (CAPEX), operating costs price; NPV decreases to 21 million US$ with an increase of 5% in (OPEX), and the discount rate (iÞ. The levels correspond to the value operating costs. The results of the sensitivity analysis of the NPV to used in the economic assessment, then low and high levels for the the capital cost show that as the investment cost increases by 5%, same value were multiplied by 0.85 and 1.15, respectively, which the NPV decreases by ca. 15 million US$. means the experimental design results are valid in the range The discount rate varied between 0.05 and 0.3. The NPV is not a between 15% and þ15%. A percentage of 15% was chosen to ensure linear function of the discount rate, the value considered was 0.1. a good adjustment. The values tested for the discount rate are 0.05, When the discount rate is 0.11, NPV decreases by approximately 21 0.1, and 0.15. ANOVA results show which parameters and in- million US$. With a discount rate higher than 0.1, NPV becomes teractions influence the NPV by analyzing the p-value. For the p- negative, making the project unviable. value < 0.01 all linear parameters and the interaction with the Results of sensitivity analysis of NPV for vanadium pentoxide discount rate were significant. Also, the statistical analysis confirms production are shown in Fig. 3. When the price increases by 5%, that price and the discount rate are the parameters exerting greater NPV increased by ca. 14 million US$. When the price drops by 26%, influence. Regression models obtained have the following form: NPV becomes negative and the project unviable. Results of the sensitivity analysis of NPV to operating costs show 2 that NPV is slightly sensitive to changes in operating costs. When NPV ¼ a þ b CAPEX þ c OPEX þ d price þ eiþ fi þ g CAPEX i operating costs increase by 5%, the NPV decreases by ca. 5 million þ h OPEX i þ j price i US$. Sensitivity analysis of the NPV to changes in capital cost shows that with an increase of 5% in the capital cost, the NPV decreases by The values for a; b; c; d; e; f ; g; h and j are 17.6, 0.9103, ca. 5 million US$. The values of NPV are very similar for both 59.55, 67.1, 1766, 14323, 0.0, 242.9, and 299.5 for REOs project, operating costs and capital costs. and 47.64, 0.9932, 12.572, 12.572, 1143.3, 5717, 0.828, The sensitivity analysis of NPV to changes in the discount rate 49.63, 49.625 for V2O5 project, respectively. The units for NPV and

Fig. 3. Sensitivity analysis, a) Sensitivity of the NPV (REOs project) to the price of REOs, operating costs and capital costs; b) Sensitivity of NPV to discount rate in REOs project; c) sensitivity of NPV (vanadium project) to the price of V2O5, operating costs and capital costs; d) Sensitivity of NPV to discount rate in vanadium project. 8 N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555

CAPEX are million US$, OPEX and price are kUS$/ton, and the discount evaluation of geochemical content, identification of suitable tech- rate is dimensionless. The R-squared values or the coefficient of the nologies, and an economic analysis will help to find more sus- 2 ¼ : 2 ¼ : 2 ¼ : regressions were R 98 17%, R adj 97 99%, and R pred 97 79% tainable alternatives to CRMs production. 2 ¼ : 2 ¼ : 2 ¼ fi for REOs project, and R 99 95%, R adj 99 95%, and R pred The DCF is a widely used method of nancial assessment, but it 2 99:94% for the V2O5 project. The R for both projects are over 98% is not a decisive metrics for making a final decision on real in- which means that at least 98% of the variation of the NPV can be vestment. In order to ensure the robustness of assessment, sensi- explained by the model. Also, excellent values of adjusted R2 and tivity analysis was performed to analyze the effect of the possible predicted R2 were observed which suggests that the number of pa- fluctuations of market prices, capital and operating costs on the rameters is the model is correct and that the model is able to produce analyzed options of CRMs production. It has been found out that high quality predictions. The ANOVA results and Pareto graphics are the discount rate and both capital and operating costs play critical included in the supplementary material. Also, supplementary mate- roles in economic decisions in different areas (Choi et al., 2018; rial gives the results of the design-of-experiment and response sur- Cisternas et al., 2014; Santander et al., 2014). face methodology for the IRR which behaves differently from the NPV. Reprocessing mine tailings will also have an impact on the environment. Due to the nature of chemical and physical processes, 4. Discussion mineral processing is water and energy intensive, some quantities of solvents and reagents are used and at the end of the process, Mine tailings are waste obtained from the processing of a rock there will still be waste that should be stored in a tailing facility. The with a view to obtain one or more products that will be refined to mining waste obtained after the reprocessing of tailings should be finally get a metal(s) that is needed. Tailings should be stored in stored in a tailing facility complying with the regulations designed facilities where they are disposed in accordance with the regula- to protect people and the environment. tions binding in each region, otherwise, the consequences to the environment can be devastating. 5. Conclusions The lack of a long-term consideration of the entire life-cycle of a mine and the instability of mine projects contribute to irreversible There are 696 tailings storage facilities in Chile, mainly from mineral losses and resource sterilization. With this knowledge in copper mining, which is the biggest mining industry in the country. mind, further research should address new strategies to anticipate The biggest TSF has the capacity to store 4,500,000,000 tons of the future use of material beyond the closing of a mine (Lebre et al., tailings. Currently, there are some initiatives for recovering metals 2017). Mine waste hierarchy goes from prevention as the most of interest from mine tailings, but such initiatives are all in the early favorable option to treatment and disposal as the least favorable stages of feasibility assessment. This study provides valuable in- options; if waste cannot be prevented then reuse and recycling are formation for the assessment of the techno-economic feasibility of needed (Lottermoser, 2011). Nowadays most mine tailings go to the industrial-scale critical materials recovery from copper industry treatment and disposal phase. In the Sustainable Development tailings. Goals, the World Economic Forum suggests the re-use of tailings, Copper production will continue to grow as the copper grade these goals are meant to be achieved by 2030 (World Economic decrease. Therefore, the volume of mine tailings that are produced Forum, 2016). The reprocessing of mine tailings is also an every year will increase as well. Mine tailings are a worldwide element of the transformation from a linear to a circular economy environmental problem as they can generate acid drainage, and that the mining industry must face. Reprocessing mine tailings to cause air pollution and soil contamination. Yet, mine tailings obtain critical materials reduces the dependency on reserve contain several valuable elements, among them critical raw mate- extraction (El Wali et al., 2019). rials. Therefore, the use of mine tailings as a secondary source Other approaches to mine tailings management from a circular would help mitigate shortages in critical raw materials by mini- economy point of view include recovering water from mine tail- mizing the reliance on primary sources. ings, which helps to reduce the reliance on seawater (Cisternas and Chilean copper mine tailings have substantial economic poten- Galvez, 2018). Recovering water or reducing the amount of water in tial as a source of critical materials such as vanadium, cobalt, rare tailing diminish the need to pump water, which decreases energy earth elements and antimony. Minerals contained in Chilean mine consumption and greenhouse gas emissions involved in pumping tailings from copper production are mostly silicates with a low water to high altitudes, where mines are usually located in Chile grade of CRMs; currently, no approved projects exist that consider (Araya et al., 2018; Herrera-Leon et al., 2019; Ramírez et al., 2019). mine tailings as a source of CRMs. Although mine tailings have a Another approach is to use mine tailings as cementitious materials low grade of CRMs, their already stored quantity is enormous. In and pigment for sustainable paints (Barros et al., 2018; Vargas and addition, prices of critical raw materials can be very high, and these Lopez, 2018). factors could make a future production of CRMs from mine tailings There have been conducted several studies on new technologies feasible. or processes to recover CRMs from secondary sources such as mine Two options of producing CRMs using mine tailings were waste (Alcalde et al., 2018; Andersson et al., 2018; Figueiredo et al., assessed; production of rare earth oxides (REOs) and production of 2018; Khalil et al., 2019; Markovaara-Koivisto et al., 2018; Peelman vanadium pentoxide (V2O5). The DFC method was used to evaluate et al., 2018). Most of these studies are carried out at laboratory and the economic feasibility of both operations. The NPV and IRR for the pilot plant scale. Nevertheless, the literature on the recovery of production of REOs are positive, which means that the project is CRMs from mine tailings is constantly growing. It is due to the fact feasible. Nevertheless, the NPV is low for an investment of this scale that new sources of CRMs are urgently needed as their importance and the IRR is close to the discount rate value. The sensitivity in the global economy is constantly growing. Moreover, the utili- analysis of the NPV of REOs production from mine tailings showed zation of wastes such as mine tailings, instead of mineral deposits, that NPV is highly sensitive to the discount rate and REO prices. is essential from a circular economy point of view. Therefore, Results of the ANOVA confirm that the discount rate and price are extrapolation of the potential of these technologies is immensely the most significant variables influencing the NPV behavior. needed. Vanadium pentoxide production is feasible for an investment of Results show that mine tailings facilities of the copper industry 14 years, as the NPV is 76 million US$ and the IRR IS 21% for V2O5 in Chile store valuable elements such as CRMs. Therefore, the early production. Vanadium is the main CRMs found in tailings in the N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555 9

Second Region in Chile. It is concluded that producing CRMs using Rare-earth economics: the balance problem. JOM (J. Occup. Med.) 65, 846e848. inactive tailings and later tailings from the active mining processes https://doi.org/10.1007/s11837-013-0639-7. Blengini, G.A., Nuss, P., Dewulf, J., Nita, V., Peiro, L.T., Vidal-Legaz, B., Latunussa, C., may be a feasible option to ensure profitable use of mine tailings Mancini, L., Blagoeva, D., Pennington, D., Pellegrini, M., Van Maercke, A., and to diversify CRMs supply. Solar, S., Grohol, M., Ciupagea, C., 2017. EU methodology for critical raw materials assessment: policy needs and proposed solutions for incremental im- provements. Resour. Pol. 53, 12e19. https://doi.org/10.1016/ Declaration of competing interest j.resourpol.2017.05.008. Ceniceros-Gomez, A.E., Macías-Macías, K.Y., de la Cruz-Moreno, J.E., Gutierrez- Ruiz, M.E., Martínez-Jardines, L.G., 2018. Characterization of mining tailings in The authors declare that they have no known competing fi Mexico for the possible recovery of strategic elements. J. South Am. Earth Sci. nancial interests or personal relationships that could have 88, 72e79. https://doi.org/10.1016/j.jsames.2018.08.013. appeared to influence the work reported in this paper. Choi, C.H., Eun, J., Cao, J., Lee, S., Zhao, F., 2018. Global strategic level supply planning of materials critical to clean energy technologies e a case study on indium. Energy 147, 950e964. https://doi.org/10.1016/j.energy.2018.01.063. CRediT authorship contribution statement Cisternas, L.A., Galvez, E.D., 2018. The use of seawater in mining. Miner. Process. Extr. Metall. Rev. 39, 18e33. https://doi.org/10.1080/08827508.2017.1389729. Natalia Araya: Conceptualization, Methodology, Validation, Cisternas, L.A., Lucay, F., Galvez, E.D., 2014. Effect of the objective function in the design of concentration plants. Miner. Eng. 63, 16e24. https://doi.org/10.1016/ Formal analysis, Investigation, Writing - original draft, Visualiza- j.mineng.2013.10.007. tion. Andrzej Kraslawski: Conceptualization, Methodology, Curry, J.A., Ismay, M.J.L., Jameson, G.J., 2014. Mine operating costs and the potential Writing - review & editing, Supervision. Luis A. Cisternas: impacts of energy and grinding. Miner. Eng. 56, 70e80. https://doi.org/10.1016/ & j.mineng.2013.10.020. Conceptualization, Validation, Formal analysis, Writing - review De Reyck, B., Degraeve, Z., Vandenborre, R., 2008. Project options valuation with net editing, Visualization, Supervision, Funding acquisition. present value and decision tree analysis. Eur. J. Oper. Res. 184, 341e355. https:// doi.org/10.1016/j.ejor.2006.07.047. De Sa, G., 2019. Brazil’s deadly dam disaster may have been preventable [WWW Acknowledgments Document]. January 29. URL. https://www.nationalgeographic.com/ environment/2019/01/brazil-brumadinho-mine-tailings-dam-disaster-could- have-been-avoided-say-environmentalists/. accessed 3.27.19. This publication was supported by Agencia Nacional de Inves- e Dold, B., Fontbote, L., 2001. Element cycling and secondary mineralogy in porphyry tigacion y Desarrollo de Chile (ANID), Anillo Grant ACM 170005. N. copper tailings as a function of climate, primary mineralogy, and mineral Araya thanks the ANID (2017, No. 21170815) for a scholarship in processing. J. Geochem. Explor. 74, 3e55. https://doi.org/10.1016/S0375- support of her doctoral studies. L.A.C. thanks the supported of 6742(01)00174-1. Edraki, M., Baumgartl, T., Manlapig, E., Bradshaw, D., Franks, D.M., Moran, C.J., 2014. MINEDUCUA project, code ANT1856 and Fondecyt program grant Designing mine tailings for better environmental, social and economic out- number 1180826. The authors are grateful to Peter Jones for his help comes: a review of alternative approaches. J. Clean. Prod. 84, 411e420. https:// in editing the paper. doi.org/10.1016/j.jclepro.2014.04.079. El Wali, M., Golroudbary, S.R., Kraslawski, A., 2019. Impact of recycling improvement on the life cycle of phosphorus. Chin. J. Chem. Eng. 27, 1219e1229. https:// Appendix A. Supplementary data doi.org/10.1016/j.cjche.2018.09.004. European Commission, 2017a. Communication from the Commission to the Euro- pean Parliament, the Council, the European Economic and Social Committee Supplementary data to this article can be found online at and the Committee of the Regions on the 2017 List of Critical Raw Materials for https://doi.org/10.1016/j.jclepro.2020.121555. the EU. European Commission, 2017b. Study on the review of the list of critical raw materials - publications Office of the EU [WWW Document]. Publ. Off. EU. URL. References https://publications.europa.eu/en/publication-detail/-/publication/08fdab5f- 9766-11e7-b92d-01aa75ed71a1/language-en. accessed 6.10.18. Abisheva, Z.S., Karshigina, Z.B., Bochevskaya, Y.G., Akcil, A., Sargelova, E.A., European Commission, 2014. Communication from the commission to the Euro- Kvyatkovskaya, M.N., Silachyov, I.Y., 2017. Recovery of rare earth metals as pean parliament, the Council, the European economic and social committee and critical raw materials from phosphorus slag of long-term storage. Hydromet- the committee of the regions on the review of the list of critical raw materials allurgy 173, 271e282. https://doi.org/10.1016/j.hydromet.2017.08.022. for the EU and the implementation of the raw materia [WWW Document]. URL. Adiansyah, J.S., Rosano, M., Vink, S., Keir, G., 2015. A framework for a sustainable https://eur-lex.europa.eu/legal-content/EN/TXT/?uri¼CELEX:52014DC0297. approach to mine tailings management: disposal strategies. J. Clean. Prod. 108, accessed 6.12.18. 1050e1062. https://doi.org/10.1016/j.jclepro.2015.07.139. European Commission, 2011. Communication from the commission to the European Ahmadi, A., Khezri, M., Abdollahzadeh, A.A., Askari, M., 2015. Bioleaching of copper, parliament, the Council, the European economic and social committee and the nickel and cobalt from the low grade sulfidic tailing of Golgohar Iron Mine, Iran. committee of the regions teckling the challenges in commodity markets and on Hydrometallurgy 154, 1e8. https://doi.org/10.1016/j.hydromet.2015.03.006. raw materials [WWW Document]. URL. https://eur-lex.europa.eu/legal-content/ Alcalde, J., Kelm, U., Vergara, D., 2018. Historical assessment of metal recovery EN/TXT/?uri¼CELEX:52011DC0025. accessed 6.3.18. potential from old mine tailings: a study case for porphyry copper tailings, Falagan, C., Grail, B.M., Johnson, D.B., 2017. New approaches for extracting and Chile. Miner. Eng. 127, 334e338. https://doi.org/10.1016/j.mineng.2018.04.022. recovering metals from mine tailings. Miner. Eng. 106, 71e78. https://doi.org/ Anderson, C.G., 2012. The metallurgy of antimony. Chem. Erde 72, 3e8. https:// 10.1016/j.mineng.2016.10.008. doi.org/10.1016/j.chemer.2012.04.001. Figueiredo, J., Vila, M.C., Matos, K., Martins, D., Futuro, A., Dinis, M. de L., Gois, J., ~ Andersson, M., Finne, T.E., Jensen, L.K., Eggen, O.A., 2018. Geochemistry of a copper Leite, A., Fiúza, A., 2018. Tailings reprocessing from Cabeço do Piao dam in mine tailings deposit in Repparfjorden, northern Norway. Sci. Total Environ. Central Portugal: a kinetic approach of experimental data. J. Sustain. Min. 17, 644, 1219e1231. https://doi.org/10.1016/j.scitotenv.2018.06.385. 139e144. https://doi.org/10.1016/j.jsm.2018.07.001. Araya, N., Lucay, F.A., Cisternas, L.A., Galvez, E.D., 2018. Design of desalinated water Foo, N., Bloch, H., Salim, R., 2018. The optimisation rule for investment in mining distribution networks: complex topography, energy production, and parallel projects. Resour. Pol. 55, 123e132. https://doi.org/10.1016/ pipelines. Ind. Eng. Chem. Res. 57 https://doi.org/10.1021/acs.iecr.7b05247. j.resourpol.2017.11.005. Barros, L., Andrade, H.D., Brigolini, G.J., Peixoto, F., Mendes, J.C., Andr, R., 2018. Reuse Ghorbani, Y., Kuan, S.H., 2017. A review of sustainable development in the Chilean of iron ore tailings from tailings dams as pigment for sustainable paints iron ore mining sector: past, present and future. Int. J. Min. Reclamat. Environ. 31, tailings dams failures in Brazil, vol. 200, pp. 412e422. https://doi.org/10.1016/ 137e165. https://doi.org/10.1080/17480930.2015.1128799. j.jclepro.2018.07.313. Graedel, T.E., Harper, E.M., Nassar, N.T., Nuss, P., Reck, B.K., Turner, B.L., 2015. Crit- Bhojwani, S., Topolski, K., Mukherjee, R., Sengupta, D., El-Halwagi, M.M., 2019. icality of metals and metalloids. Proc. Natl. Acad. Sci. U.S.A. 112, 4257e4262. Technology review and data analysis for cost assessment of water treatment https://doi.org/10.1073/pnas.1500415112. systems. Sci. Total Environ. https://doi.org/10.1016/j.scitotenv.2018.09.363. Herrera-Leon, S., Lucay, F.A., Cisternas, L.A., Kraslawski, A., 2019. Applying a multi- Binnemans, K., Jones, P.T., Blanpain, B., Van Gerven, T., Pontikes, Y., 2015. Towards objective optimization approach in designing water supply systems for mining zero-waste valorisation of rare-earth-containing industrial process residues: a industries. The case of Chile. J. Clean. Prod. 210 https://doi.org/10.1016/ critical review. J. Clean. Prod. 99, 17e38. https://doi.org/10.1016/ j.jclepro.2018.11.081. j.jclepro.2015.02.089. Hornby, L., Sanderson, H., 2019. Rare earths: beijing threatens a new front in the Binnemans, Koen, Jones, P.T., Blanpain, B., Van Gerven, T., Yang, Y., Walton, A., trade war [WWW Document]. Financ. Times. URL. https://www.ft.com/content/ Buchert, M., 2013a. Recycling of rare earths: a critical review. J. Clean. Prod. 51, 3cd18372-85e0-11e9-a028-86cea8523dc2. accessed 2.5.20. 1e22. https://doi.org/10.1016/j.jclepro.2012.12.037. Hudson Resources Inc, 2013. Sarfartoq rare earth elements project [WWW Docu- Binnemans, K., Jones, P.T., Van Acker, K., Blanpain, B., Mishra, B., Apelian, D., 2013b. ment]. URL. https://hudsonresourcesinc.com/projects/sarfartoq-rare-earth- 10 N. Araya et al. / Journal of Cleaner Production 263 (2020) 121555

element-project/. accessed 6.6.18. 2019. The potential of iron ore tailings as secondary deposits of rare earths. Ibanez-For~ es, V., Bovea, M.D., Perez-Belis, V., 2014. A holistic review of applied Appl. Sci. 9, 2913. https://doi.org/10.3390/app9142913. methodologies for assessing and selecting the optimal technological alternative Mudd, G.M., Jowitt, S.M., Werner, T.T., 2018. Global platinum group element re- from a sustainability perspective. J. Clean. Prod. 70, 259e281. https://doi.org/ sources, reserves and mining e a critical assessment. Sci. Total Environ. 10.1016/j.jclepro.2014.01.082. 622e623, 614e625. https://doi.org/10.1016/j.scitotenv.2017.11.350. Innocenzi, V., De Michelis, I., Kopacek, B., Veglio, F., 2014. Yttrium recovery from Onishi, K., Nakamura, T., Nishihama, S., Yoshizuka, K., 2010. Synergistic solvent primary and secondary sources: a review of main hydrometallurgical processes. impregnated resin for adsorptive separation of lithium ion. Ind. Eng. Chem. Res. Waste Manag. 34, 1237e1250. https://doi.org/10.1016/j.wasman.2014.02.010. 49, 6554e6558. https://doi.org/10.1021/ie100145d. Jordens, A., Cheng, Y.P., Waters, K.E., 2013. A review of the beneficiation of rare earth Onyedika, G.O., Achusim-Udenko, A.C., Nwoko, C.I.A., Ogwuegbu, M.O.C., 2012. element bearing minerals. Miner. Eng. 41, 97e114. https://doi.org/10.1016/ Chemistry, processes and problems of complex ores utilization: hydrometal- j.mineng.2012.10.017. lurgical options. Int. J. Chem. Sci. 10, 112e130. Jorjani, E., Shahbazi, M., 2016. The production of rare earth elements group via Oyarzún, J., Oyarzun, R., Lillo, J., Higueras, P., Maturana, H., Oyarzún, R., 2016. Dis- tributyl phosphate extraction and precipitation stripping using oxalic acid. tribution of chemical elements in calc-alkaline igneous rocks, soils, sediments Arab. J. Chem. 9, S1532eS1539. https://doi.org/10.1016/j.arabjc.2012.04.002. and tailings deposits in northern central Chile. J. South Am. Earth Sci. 69, Khalil, A., Argane, R., Benzaazoua, M., Bouzahzah, H., Taha, Y., Hakkou, R., 2019. 25e42. https://doi.org/10.1016/j.jsames.2016.03.004. PbeZn mine tailings reprocessing using centrifugal dense media separation. Peelman, S., Kooijman, D., Sietsma, J., Yang, Y., 2018. Hydrometallurgical recovery of Miner. Eng. 131, 28e37. https://doi.org/10.1016/j.mineng.2018.10.023. rare earth elements from mine tailings and WEEE. J. Sustain. Metall. 4, 367e377. Khorasanipour, M., 2015. Environmental mineralogy of Cu-porphyry mine tailings, a https://doi.org/10.1007/s40831-018-0178-0. case study of semi-arid climate conditions, sarcheshmeh mine, SE Iran. Peelman, S., Sun, Z.H., Sietsma, J., Yang, Y., 2016. Hydrometallurgical extraction of J. Geochem. Explor. 153, 40e52. https://doi.org/10.1016/j.gexplo.2015.03.001. rare earth elements from low grade mine tailings. Rare Met. Technol. 2016, Khorasanipour, M., Jafari, Z., 2017. Environmental geochemistry of rare earth ele- 17e29. https://doi.org/10.1007/978-3-319-48135-7_2. ments in Cu-porphyry mine tailings in the semiarid climate conditions of Peelman, S., Sun, Z.H.I., Sietsma, J., Yang, Y., 2014. Leaching of rare earth Elements : Sarcheshmeh mine in southeastern Iran. Chem. Geol. 477, 58e72. https:// past and present. In: ERES2014 1st Eur. Rare Earth Resour. Conf., pp. 446e456. doi.org/10.1016/j.chemgeo.2017.12.005. https://doi.org/10.1016/B978-0-12-802328-0.00021-8. Kinnunen, P.H.M., Kaksonen, A.H., 2019. Towards circular economy in mining: op- Ramírez, Y., Kraslawski, A., Cisternas, L.A., 2019. Decision-support framework for the portunities and bottlenecks for tailings valorization. J. Clean. Prod. 228, environmental assessment of water treatment systems. J. Clean. Prod. 225, 153e160. https://doi.org/10.1016/j.jclepro.2019.04.171. 599e609. https://doi.org/10.1016/j.jclepro.2019.03.319. Kirchherr, J., Reike, D., Hekkert, M., 2017. Conceptualizing the circular economy: an Sadri, F., Nazari, A.M., Ghahreman, A., 2017. A review on the cracking, baking and analysis of 114 definitions. Resour. Conserv. Recycl. 127, 221e232. https:// leaching processes of rare earth element concentrates. J. Rare Earths 35, doi.org/10.1016/j.resconrec.2017.09.005. 739e752. https://doi.org/10.1016/S1002-0721(17)60971-2. Kodukula, P., Papudesu, C., 2006. Project Valuation Using Real Options - A Practi- Santander, C., Robles, P.A., Cisternas, L.A., Rivas, M., 2014. Technical-economic tioner’s Guide. J. Ross Publishing, Fort Lauderdale, Florida. feasibility study of the installation of biodiesel from microalgae crops in the Koltun, P., Tharumarajah, A., 2014. Life cycle impact of rare earth elements. ISRN Atacama Desert of Chile. Fuel Process. Technol. 125, 267e276. https://doi.org/ Metall. 1e10 https://doi.org/10.1155/2014/907536,2014. 10.1016/j.fuproc.2014.03.038. Kossoff, D., Dubbin, W.E., Alfredsson, M., Edwards, S.J., Macklin, M.G., Hudson- Schoenberger, E., 2016. Environmentally sustainable mining: the case of tailings Edwards, K.A., 2014. Mine tailings dams: characteristics, failure, environmental storage facilities. Resour. Pol. 49, 119e128. https://doi.org/10.1016/ impacts, and remediation. Appl. Geochem. 51, 229e245. https://doi.org/ j.resourpol.2016.04.009. 10.1016/j.apgeochem.2014.09.010. SERNAGEOMIN, 2018. Datos públicos deposito de Relaves [WWW Document]. URL. Larsson, M., Nosrati, A., Kaur, S., Wagner, J., Baus, U., Nyden, M., 2018. Copper https://www.sernageomin.cl/datos-publicos-deposito-de-relaves/. accessed removal from acid mine drainage-polluted water using glutaraldehyde- 5.10.18. polyethyleneimine modified diatomaceous earth particles. Heliyon 4, e00520. SERNAGEOMIN, 2017. Anuario de la Minería de Chile [WWW Document]. URL. https://doi.org/10.1016/j.heliyon.2018.e00520. https://www.sernageomin.cl/anuario-de-la-mineria-de-chile/. accessed 9.10.18. Leader, A., Gaustad, G., Babbitt, C., 2019. The effect of critical material prices on the SERNAGEOMIN, 2011. Ley 20.551 - cierre de Faenas mineras [WWW Document]. competitiveness of clean energy technologies. Mater. Renew. Sustain. Energy 8, URL. https://www.sernageomin.cl/cierre-de-faenas-mineras/. accessed 2.5.20. 1e17. https://doi.org/10.1007/s40243-019-0146-z. Sracek, O., Mihaljevic, M., Kríbek, B., Majer, V., Veselovský, F., 2010. Geochemistry Lebre, E., Corder, G., Golev, A., 2017. The role of the mining industry in a circular and mineralogy of Cu and Co in mine tailings at Copperbelt, Zambia. J. Afr. Earth economy: a framework for resource management at the mine site level. J. Ind. Sci. 57, 14e30. https://doi.org/10.1016/j.jafrearsci.2009.07.008. Ecol. 21, 662e672. https://doi.org/10.1111/jiec.12596. Sun, X., Ji, Y., Chen, J., Ma, J., 2009. Solvent impregnated resin prepared using task- Lee, J., 2018. Prophecy announces positive preliminary economic assessment study specific ionic liquids for rare earth separation. J. Rare Earths 27, 932e936. for the Gibellini vanadium project - junior mining network [WWW Document]. https://doi.org/10.1016/S1002-0721(08)60365-8. URL. https://www.juniorminingnetwork.com/junior- miner-news/press- Tapia, J., Davenport, J., Townley, B., Dorador, C., Schneider, B., Tolorza, V., von releases/693-tsx/pcy/47430-prophecy-announces-positive-preliminary-eco- Tümpling, W., 2018. Sources, enrichment, and redistribution of As, Cd, Cu, Li, nomic-assessment-study-for-the-gibellini-vanadium-project.html. accessed Mo, and Sb in the Northern Atacama Region, Chile: implications for arid wa- 6.5.18. tersheds affected by mining. J. Geochem. Explor. 185, 33e51. https://doi.org/ LME, 2018. London metal exchange: LME cobalt [WWW Document]. URL. https:// 10.1016/j.gexplo.2017.10.021. www.lme.com/en-GB/Metals/Minor-metals/Cobalt#tabIndex¼0. accessed Thenorthernminer.com, 2018. The northern miner e mining news since 1915 1.7.18. [WWW Document]. URL. https://www.northernminer.com/. accessed 1.7.18. Lottermoser, B.G., 2011. Recycling, reuse and rehabilitation of mine wastes. Ele- Tunsu, C., Menard, Y., Eriksen, D.Ø., Ekberg, C., Petranikova, M., 2019. Recovery of ments 7, 405e410. https://doi.org/10.2113/gselements.7.6.405. critical materials from mine tailings: a comparative study of the solvent Markovaara-Koivisto, M., Valjus, T., Tarvainen, T., Huotari, T., Lerssi, J., Eklund, M., extraction of rare earths using acidic, solvating and mixed extractant systems. 2018. Preliminary volume and concentration estimation of the Aijala tailings J. Clean. Prod. 218, 425e437. https://doi.org/10.1016/j.jclepro.2019.01.312. pond e evaluation of geophysical methods. Resour. Pol. 59, 7e16. https:// Vargas, F., Lopez, M., 2018. Development of a new supplementary cementitious doi.org/10.1016/j.resourpol.2018.08.016. material from the activation of copper tailings: mechanical performance and Mendez, M.O., Maier, R.M., 2008. Phytostabilization of mine tailings in arid and analysis of factors. J. Clean. Prod. 182, 427e436. https://doi.org/10.1016/ semiarid environments - an emerging remediation technology. Environ. Health j.jclepro.2018.01.223. Perspect. 116, 278e283. https://doi.org/10.1289/ehp.10608. Vekasi, K., Hunnewell, N.L., 2019. China’s control of rare earth metals [WWW Mineralprices.com, 2018. Rare earth metals [WWW Document]. URL. http:// Document]. Natl. Bur. Asian Res. URL. https://www.nbr.org/publication/chinas- mineralprices.com/rare-earth-metals/. accessed 7.3.18. control-of-rare-earth-metals/. accessed 2.5.20. Ministerio de Minería, 2011. Ley 20.551 - ministerio de Minería [WWW Document]. World Economic Forum, 2016. Mapping mining to the sustainable development URL. http://www.minmineria.gob.cl/leyes-sectoriales/ley-20551/. accessed goals: an atlas [WWW Document]. URL. https://www.undp.org/content/dam/ 2.5.20. undp/library/Sustainable Development/Extractives/Mapping_Mining_SDGs_ Miyauchi, A., Okabe, T.H., 2010. Production of metallic vanadium by preform An_Atlas_Executive_Summary_FINAL.pdf. accessed 2.7.20. reduction process. Mater. Trans. 51, 1102e1108. https://doi.org/10.2320/ Xiang, J., Huang, Q., Lv, X., Bai, C., 2018. Extraction of vanadium from converter slag matertrans.M2010027. by two-step sulfuric acid leaching process. J. Clean. Prod. 170, 1089e1101. Mohamed, S., van der Merwe, E.M., Altermann, W., Doucet, F.J., 2017. Process https://doi.org/10.1016/j.jclepro.2017.09.255. development for elemental recovery from PGM tailings by thermochemical Yoon, H.S., Kim, C.J., Chung, K.W., Kim, S.D., Lee, J.Y., Kumar, J.R., 2016. Solvent treatment: preliminary major element extraction studies using ammonium extraction, separation and recovery of dysprosium (Dy) and neodymium (Nd) sulphate as extracting agent. Waste Manag. 66, 222e224. https://doi.org/ from aqueous solutions: waste recycling strategies for permanent magnet 10.1016/j.wasman.2017.04.009. processing. Hydrometallurgy 165, 27e43. https://doi.org/10.1016/ Moodley, I., Sheridan, C.M., Kappelmeyer, U., Akcil, A., 2017. Environmentally sus- j.hydromet.2016.01.028. tainable acid mine drainage remediation: research developments with a focus Zizlavský, O., 2014. Net present value approach: method for economic assessment on waste/by-products. Miner. Eng. 126, 1e14. https://doi.org/10.1016/ of innovation projects. Procedia - Soc. Behav. Sci. 156, 506e512. https://doi.org/ j.mineng.2017.08.008. 10.1016/j.sbspro.2014.11.230. Moran-Palacios, H., Ortega-Fernandez, F., Lopez-Castano,~ R., Alvarez-Cabal, J.V.,

PUBLICATION III

Araya, N., Ramírez, Y., Kraslawski, A., Cisternas, L.A.

Feasibility of re-processing mine tailings to obtain critical materials using real options analysis (submitted)

Submitted in Sustainable Development

Sustainable Development

Feasibility of re-processing mine tailings to obtain critical materials using real options analysis

Journal: Sustainable Development

Manuscript ID SD-20-0738 Wiley - Manuscript type:ForResearch Peer Article Review critical materials, real options, circular economy, mine tailings Keywords: valorization, mine tailings management, net present value

http://mc.manuscriptcentral.com/sd Page 1 of 49 Sustainable Development

1 2 3 4 Feasibility of re-processing mine tailings to obtain critical materials using real 5 6 options analysis 7 8 Abstract 9 10 The re-processing of mine tailings to obtain critical materials could reduce the mining of new 11 12 13 deposits as well as ensure the profitable use of the waste materials. Though, it requires 14 15 large scale industrial installations and the development of specialized technologies to obtain 16 17 critical materials. New investment in mining activities is an operation, engaging for 18 19 considerable financial Forresources Peer involved. TheReview scale of such an endeavor makes a new 20 21 mining activity a high-risk operation due to several uncertainties present. Therefore, there is 22 23 an acute need to use new tools to assess the risk associated with the planning and 24 25 development of new mining activities. 26 27 28 This study introduces a framework to evaluate the economic risk related to the re-processing 29 30 of mine tailings to obtain critical materials. The framework, based on real options analysis 31 32 (ROA), and sensitivity and uncertainty analysis, was applied to analyze the profitability of 33 34 using mine tailings as a source of critical materials in the Chilean mining industry. 35 36 37 Results show that tailing storage facilities in Chile have some stocks of critical materials, like 38 39 scandium, whose extraction could be profitable. For the data used, the results of uncertainty 40 41 and sensitivity analyses show that capital expenditure has a more significant influence than 42 43 the other variables. Therefore, for the case of mine tailings re-processing, it is essential to 44 45 develop processes and technologies that enable lower capital expenses. 46 47 48 Keywords: Critical Materials; Real Options; Circular Economy; Mine Tailings Valorization; Mine 49 50 Tailings Management; Net Present Value 51 52 53 54 55 56 57 58 1 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 2 of 49

1 2 3 1. Introduction 4 5 Critical materials (CMs) are vital for the functioning of modern society. They are fundamental 6 7 to the production of a broad range of goods and applications in our everyday life. Some raw 8 9 materials are more important than others due to their role in digital technologies, low-carbon 10 11 12 technologies, and sustainable mobility (David & Koch, 2019; Mathieux et al., 2017; X. Wang 13 14 et al., 2017). According to European Commission, critical raw materials are a group of raw 15 16 materials possessing two common characteristics: they have a high economic importance 17 18 to European Union and their supply is associated with high risk (European Commission, 19 For Peer Review 20 2017b). The recent list made by European Commission includes elements such as rare 21 22 Earth Elements (REEs), vanadium, cobalt, platinum group metals, among others (European 23 24 Commission, 2017b). 25 26 27 Currently, extracting raw materials and metals from ore bodies of declining ore grade means 28 29 potentially bigger mine size generating more waste, overburden to the environment, higher 30 31 consumption of energy, water, and auxiliary materials which overall produce severe 32 33 environmental consequences and larger political risks (de Koning et al., 2018; Mudd, 2010; 34 35 Northey, Mudd, Saarivuori, Wessman-Jääskeläinen, & Haque, 2016; Žibret et al., 2020). 36 37 38 Furthermore, by 2050 the overall demand for metals will rise by a factor of 3-4 (de Koning 39 40 et al., 2018). When a mine has exhausted its resources whose extraction is economically 41 42 viable, an alternative is to close it and then reopen when the market and technology 43 44 conditions enable a profitable re-processing of tailings (Northey et al., 2016). 45 46 47 Long-term supply of metals highly depends on actual and expected prices, cumulative 48 49 availability curve, and new mining technologies (de Koning et al., 2018; Gordon, Bertram, & 50 51 Graedel, 2007; Tilton & Lagos, 2007; Yaksic & Tilton, 2009). Additionally, the development 52 53 of new high-tech technologies might lead to disruptive demand change (Tukker, 2014); 54 55 56 57 58 2 59 60 http://mc.manuscriptcentral.com/sd Page 3 of 49 Sustainable Development

1 2 3 meanwhile, it may take ten years or more to open new mines and adjust production (de 4 5 Koning et al., 2018). 6 7 8 Mine tailings are waste obtained after the processing of some minerals to acquire one or 9 10 more elements of interest. They are composed of a mixture of heavy metals, water, sand, 11 12 and fine-grained solid material, and generally are deposited in ponds without further 13 14 treatment (Babel, Chauhan, Ali, & Yadav, 2016; Santibañez et al., 2012; L. Wang, Ji, Hu, 15 16 Liu, & Sun, 2017). 17 18 19 Mine tailings deposits Formay contain Peer many valuable Review elements as has been shown in several 20 21 studies (Alcalde, Kelm, & Vergara, 2018; Andersson, Finne, Jensen, & Eggen, 2018; 22 23 Figueiredo et al., 2018; Khalil et al., 2019; Khorasanipour & Jafari, 2017; Macías-Macías, 24 25 26 Ceniceros-Gómez, Gutiérrez-Ruiz, González-Chávez, & Martínez-Jardines, 2019; 27 28 Markovaara-Koivisto et al., 2018; Medas, Cidu, De Giudici, & Podda, 2013). In the particular 29 30 case of Chilean copper mine tailings, 0.82% are metals, non-metals, and metalloids, 0.01% 31 32 rare earth elements, and the rest major rock-forming elements (SERNAGEOMIN, 2017). 33 34 Nowadays, mine tailings are increasingly often seen as a potential source of raw materials, 35 36 and secondary sources are attracting more and more attention due to the benefits of a 37 38 circular economy approach (Jeswiet, 2017; Kinnunen & Kaksonen, 2019). 39 40 41 Mine tailings could be re-processed to extract CMs. This study aims to propose a framework 42 43 to assess the feasibility of re-processing mine tailings to obtain CMs. The presented 44 45 methodology is illustrated by the set of unique data applicable to Chilean mine tailings. 46 47 However, thanks to its generality, the proposed approach could be applied in other mine 48 49 tailings deposits where geochemical data are available. For example, the contents of CMs 50 51 in mine tailings have been already measured in some deposits in Finland, Sweden, Portugal, 52 53 54 Indonesia, and Mexico (Ceniceros-Gómez, Macías-Macías, de la Cruz-Moreno, Gutiérrez- 55 56 Ruiz, & Martínez-Jardines, 2018; Hällström, Alakangas, & Martinsson, 2018; Markovaara- 57 58 3 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 4 of 49

1 2 3 Koivisto et al., 2018; Peelman, Kooijman, Sietsma, & Yang, 2018; Szamałek, Konopka, 4 5 Zglinicki, & Marciniak-Maliszewska, 2013; Tunsu, Menard, Eriksen, Ekberg, & Petranikova, 6 7 2019). 8 9 10 The research questions that this study seeks to answer are: 11 12 13 • Is it economically feasible to invest in a project developed around the idea of re- 14 15 processing mine tailings to obtain critical materials? 16 17 • How can real options analysis improve the decision to invest in a project using mine 18 19 tailings as a sourceFor of critical Peer materials? Review 20 21 • How can the net present value variables influence real options performance? 22 23 24 The framework developed in this study is applied to a case study based on active mine 25 26 tailings located in northern Chile. This area features several mining projects based on 27 28 copper resources, leaving tremendous volumes of tailings that are stored without any 29 30 economical purpose. 31 32 33 The novel contribution of this article consists in considering ROA and sensitivity and 34 35 uncertainty analysis to give flexibility to the economic assessment of mine tailings re- 36 37 processing, acknowledging the uncertainties involved. Net present value (NPV) is calculated 38 39 40 using the DCF method as a starting point, then with ROA, different outcomes for an 41 42 investment project are estimated (Arnold, 2014; Brandão, Dyer, & Hahn, 2005; Kodukula & 43 44 Papudesu, 2006; Trigeorgis, 1996). The viability of mining investments depends on 45 46 variables that have high uncertainty as to market prices of metals. In the case of re- 47 48 processing mine tailings, technology developments and business model development are 49 50 still needed. To study the influence of several variables on NPV outcome, Monte Carlo 51 52 simulation is used to perform sensitivity and uncertainty analysis. 53 54 55 56 57 58 4 59 60 http://mc.manuscriptcentral.com/sd Page 5 of 49 Sustainable Development

1 2 3 1.1. The current state of mine tailings re-processing technologies 4 5 The annual amount of mine tailings generated by the mining industry exceeds 10 billion tons 6 7 (Adiansyah, Rosano, Vink, & Keir, 2015) and is expected to be growing because of the 8 9 increasing production forecast by 2035 the volume of tailings will double (CESCO, 2019). 10 11 12 Due to higher demand for mineral products and lower grades of ore as a result of which 13 14 more materials will have to be processed in more energy-intensive processes (C. Wang, 15 16 Harbottle, Liu, & Xu, 2014). Particularly, Chile has 6,500 million m3 of mine tailings and has 17 18 an approved capacity of over 14,470 million m3 (SERNAGEOMIN, 2019a). 19 For Peer Review 20 21 Mine tailings contain several CMs, even if their content is low, the volumes of mine tailings 22 23 are big enough to consider re-processing (Binnemans, Jones, Blanpain, Van Gerven, & 24 25 Pontikes, 2015; Careddu, Dino, Danielsen, & Přikryl, 2018). Several studies analyze the 26 27 geochemical content of mine tailings to demonstrate that they contain CMs and point out 28 29 that tailings could be re-processed in the future (Ceniceros-Gómez et al., 2018; Dino, Mehta, 30 31 Rossetti, Ajmone-Marsan, & De Luca, 2018; Markovaara-Koivisto et al., 2018; Moran- 32 33 Palacios, Ortega-Fernandez, Lopez-Castaño, & Alvarez-Cabal, 2019; Tunsu et al., 2019). 34 35 36 Rare earth elements (REEs) are considered CMs due to their importance for generation of 37 38 clean energy and technologies essential for the transition to a low carbon economy. The 39 40 dominant position of China as the main producer, strong dependency of industrialized 41 42 countries on imports, geopolitical factors, and high specificity are additional factors 43 44 contributing to their criticality (European Commission, 2017a; Tunsu et al., 2019). The 45 46 47 world’s REEs reserves are found in the Inner Mongolia region of North China, with 59.3% of 48 49 world amount (Zhang, Liu, Li, & Jiang, 2014). 50 51 The prospect of using waste like metal scrap or mine tailings for obtaining REEs is gaining 52 53 54 attention since it reduces the dependency on primary sources. Several studies are drawing 55 56 attention to the possibility of re-processing mine tailings to get REEs using hydrometallurgy 57 58 5 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 6 of 49

1 2 3 processes (Borra, Pontikes, Binnemans, & Van Gerven, 2015; Peelman et al., 2018; 4 5 Peelman, Sun, Sietsma, & Yang, 2016; Zhang et al., 2014). 6 7 8 Sustainable development components involved in mine tailings management strategy 9 10 should include energy, water, technology, environmental impact, cost (Adiansyah et al., 11 12 2015), and policy. The valorization of mine tailings appears a critical component to achieving 13 14 a circular economy model in the mining industry, which needs to improve its processes to 15 16 minimize the environmental impacts of mining waste (Lèbre, Corder, & Golev, 2017; Tayebi- 17 18 Khorami, Edraki, Corder, & Golev, 2019). The valorization of tailings is still in the early 19 For Peer Review 20 21 stages, but it is expected to improve in the future (Kinnunen & Kaksonen, 2019). 22 23 1.2. Project valuation tools 24 25 26 Traditional valuation tools, such as the DCF method, are static methods that do not consider 27 28 the uncertainty of the variables used to estimate the profitability of an investment (Arnold, 29 30 2014; Brandão et al., 2005; Kodukula & Papudesu, 2006; Trigeorgis, 1996). 31 32 33 Investments in the mining sector are capital intensive and mostly irrevocable with a limited 34 35 economic life span. Their economic viability depends on the uncertain world market price 36 37 and on how project risks emerge (Guj & Chandra, 2019). The project value is determined by 38 39 the commodity market and flexibility inherent in the metal mining system to respond to 40 41 uncertainties (Savolainen, Collan, & Luukka, 2017). The available future metal prices can 42 43 be used as certain in the project valuation process, where their maturity is maximum 44 45 between two to five years (Savolainen, 2016). 46 47 48 49 Project valuation techniques that ignore the real option nature of the project are widely used 50 51 in the mining industry, e.g., NPV, IRR, and static DCF (Savolainen, 2016). However, static 52 53 NPV is used as a benchmark for a real NPV option and option agreement methods (Arnold, 54 55 2014). DCF method treats future cash flows as deterministic values. Methods to include 56 57 58 6 59 60 http://mc.manuscriptcentral.com/sd Page 7 of 49 Sustainable Development

1 2 3 uncertainty in the result of NPV are: increasing the discount rate, applying sensitivity 4 5 analysis, comparing pessimistic and optimistic cash flows, or to use scenario planning to 6 7 estimate expected cash flows (Gaspars-Wieloch, 2019). 8 9 10 There are many approaches to analyze cost-intensive investments under the conditions of 11 12 uncertainty (Cristóbal et al., 2013). One of the methods gaining popularity is ROA as it is 13 14 being applied in different fields (Insley, 2002; Nelson, Howden, & Hayman, 2013; Regan et 15 16 al., 2015; Schatzki, 2003; Slade, 2001). Real options are a right, not an obligation, to 17 18 undertake business initiatives connected with tangible assets (Kodukula & Papudesu, 2006; 19 For Peer Review 20 21 T. Wang & Neufville, 2005). Real options acknowledge managerial flexibility and readiness 22 23 to adjust investment projects due to future uncertainty and the changing environment, 24 25 involving possible managerial options that can reshape a project and adapt it to changing 26 27 conditions to maintain or enhance its profitability (Trigeorgis, 1993). Real options for a 28 29 project exploit the flexibility of sequential investment with flexible strategies and the 30 31 capability to delay decisions in an engineering system to react to an uncertain outcome; 32 33 where the changes depend on exogenous and endogenous uncertainties, so it is hard to 34 35 make credible value estimates (Guj & Chandra, 2019). Common attributes of real options 36 37 38 valuation design include identification of sources of uncertainty, available real options 39 40 recognition, modeling of uncertain variables, and real option valuation (Kozlova, 2017). 41 42 Methods used to evaluate real options are decision trees, Monte Carlo simulations, and the 43 44 Black-Scholes model (Arnold, 2014; Collan, 2011; Kodukula & Papudesu, 2006). 45 46 47 ROA is of rising importance in metal mining to assess possible future metal scenarios to 48 49 cope with mining projects under growing market uncertainty and project complexity. The real 50 51 options valuation method aims to protect and increase the economic return from a project 52 53 (Savolainen, 2016) and is applied predominantly in investment project valuation (Kozlova, 54 55 2017). 56 57 58 7 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 8 of 49

1 2 3 1.3. Sensitivity and Uncertainty Analysis 4 5 A project is based on many inputs that are uncertain, such as production costs, price of 6 7 materials and equipment, and sales volume. The analysis and modeling of uncertainty 8 9 enhances the ability to make appropriate decisions. The need to analyze uncertainties 10 11 12 comes from the awareness that data abundance does not necessarily provide certitude, and 13 14 sometimes can lead to errors in the decision-making process (Attoh-Okine & Ayyub, 2005). 15 16 Due to uncertainty, there is no project without risk, and the uncertainty could be caused by 17 18 the different factors, e.g. lack of information or data (Munier, 2014). 19 For Peer Review 20 21 Sensitivity analysis examines the response or reaction of an output variable, such as the 22 23 NPV to the variations of input variables, such as price or sales volume, etc. (Munier, 2014). 24 25 Sensitivity analysis methods explore and quantify the impact of possible errors in the input 26 27 data on predicted model outputs (Loucks & van Beek, 2005). Sensitivity analysis is used in 28 29 a broad spectrum of disciplines to study how various sources of uncertainty in a model 30 31 contribute to the model’s overall uncertainty. On the other hand, uncertainty analysis 32 33 assesses the impact of ambiguous values of parameters on the final results (Cacuci, 2003). 34 35 Sensitivity analysis may be performed together with uncertainty analysis to ensure the 36 37 38 quality of the model and the transparency of the decision-making process (Borgonovo, 39 40 2017). 41 42 Monte Carlo simulation methods are tools widely used to perform sensitivity and uncertainty 43 44 analysis (Attoh-Okine & Ayyub, 2005). Monte Carlo simulation can be used for risk analysis 45 46 47 by modeling the probability of the different outcomes of a process; it can be used as a 48 49 valuation tool for projects (Arnold, 2014; Collan, 2011; Kodukula & Papudesu, 2006). Monte 50 51 Carlo simulation randomly generates values of the uncertain variables within a certain range 52 53 to simulate the potential outcomes (Kodukula & Papudesu, 2006; Mun, 2002). Sensitivity 54 55 and uncertainty analysis can be used as complementary tool to the DCF method and ROA 56 57 58 8 59 60 http://mc.manuscriptcentral.com/sd Page 9 of 49 Sustainable Development

1 2 3 to study the effect of the inputs on NPV (De Reyck, Degraeve, & Vandenborre, 2008; 4 5 Gaspars-Wieloch, 2019; Kodukula & Papudesu, 2006; Pivorienė, 2017). 6 7 8 2. Methodology 9 10 The profitability analysis of mine tailings processing requires the identification of 11 12 geochemical characteristics of the tailing deposit, also named tailing storage facility (TSF). 13 14 Mineral characteristics and concentration of elements that are present in the TSF along with 15 16 the mass of tailings accumulated in the deposit allow estimating the quantity of each CM 17 18 (Araya, Kraslawski, & Cisternas, 2020; Markovaara-Koivisto et al., 2018; Moran-Palacios et 19 For Peer Review 20 21 al., 2019). 22 23 24 In the economic analysis, in-situ value is estimated by multiplying the total mass of a specific 25 26 CM by its price. Total mass is calculated based on the average concentration and mass in 27 28 a TSF (Araya et al., 2020; Markovaara-Koivisto et al., 2018). DCF method is used to 29 30 estimate the NPV of a project for extracting one or more CMs from one of the TSF analyzed. 31 32 Criteria used to choose which CM could be extracted include concentration, total mass, and 33 34 price. In some cases, finding such data, as well as get access to data for estimating CAPEX 35 36 and OPEX for each CM is rather difficult. 37 38 39 Subsequently, an overview of technologies suitable for re-processing mine waste stored in 40 41 a TSF to extract the CMs is needed. If these technologies are still not applicable on an 42 43 industrial scale, then, technologies used to process primary ores to obtain such CM are 44 45 considered, the same goes for economic inputs such as CAPEX, OPEX, and production. 46 47 48 When dealing with mine waste, some consideration needs to be given to aspects, such as 49 50 a lower grade of elements contained in it (Araya et al., 2020; Binnemans et al., 2015; 51 52 Falagán, Grail, & Johnson, 2017). Mine tailings are already in the form of a paste or slurry, 53 54 so there are no mining costs, which usually represent 43% of operating costs in a mine 55 56 (Curry, Ismay, & Jameson, 2014). 57 58 9 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 10 of 49

1 2 3 Sensitivity and uncertainty analysis is applied to the NPV estimated by the DCF method. 4 5 6 ROA is applied using the NPV obtained with the DCF method using the method of risk- 7 8 neutral probabilities presented in. The binomial tree analysis is used to apply ROA. A 9 10 binomial tree is built by using the method of risk-neutral probabilities presented by Kodukula 11 12 13 and Papudesu (2006); this methodology involves adjusting the risk of the cash flows across 14 15 the lattice with risk-neutral probabilities and discounting them at a risk-free rate (Kodukula 16 17 & Papudesu, 2006). 18 19 For Peer Review 20 2.1. Discounted Cash Flow (DCF) method 21 22 23 DCF method is based on the calculation of NPV of a project over its entire life cycle 24 25 accounting for the investment and the free cash flows throughout its whole life (Kodukula & 26 27 Papudesu, 2006). 28 29 30 푃푟표푗푒푐푡 푁푃푉 = 푃푉 표푓 푓푟푒푒 푐푎푠ℎ 푓푙표푤푠 푖푛 푝푟표푑푢푐푡푖표푛 푝ℎ푎푠푒 ― 푃푉 표푓 푖푛푣푒푠푡푚푒푛푡푠 푐표푠푡푠 (1) 31 32 33 According to the DCF method, if the project NPV is greater than zero, it means that the 34 35 project revenues are greater than the costs of the project, so it is financially attractive 36 37 (Arnold, 2014; Kodukula & Papudesu, 2006). 38 39 40 Present value (PV) is the estimation of costs and net revenues of the development 41 42 43 production phase. These are the free cash flows over the entire life cycle of the project 44 45 (Kodukula & Papudesu, 2006). 46 47 퐹푉 48 푃푉 = (2) 49 (1 + 푟)푛 50 51 52 Where FV is the future value, r is the discount rate per time period, and n is the number of 53 54 the time period. 55 56 57 58 10 59 60 http://mc.manuscriptcentral.com/sd Page 11 of 49 Sustainable Development

1 2 3 To estimate NPV of a project, capital expenses (CAPEX) and operating expenses (OPEX) 4 5 of a project with similar characteristics in terms of ore grade and production capacity are 6 7 used. CAPEX may include treatment equipment, water intake structure, site preparation, 8 9 concentrate discharge system, auxiliary equipment, piping, valves, among other cost items 10 11 12 (Bhojwani, Topolski, Mukherjee, Sengupta, & El-Halwagi, 2019). OPEX include labor, 13 14 energy cost, chemicals, maintenance, spare parts, and indirect cost (Bhojwani et al., 2019). 15 16 17 The following equation was employed to estimate NPV and PV of a project based on the 18 19 use of mine tailings: For Peer Review 20 21 22 퐶퐴푃퐸푋 퐶퐴푃퐸푋 푐푎푠ℎ 푓푙표푤 = 푝푟푖푐푒 ∙ 푝푟표푑푢푐푡푖표푛 ― 푂푃퐸푋 ∙ 푝푟표푑푢푐푡푖표푛 ― ∙ (1 ― 푡푎푥푒푠) + (3) 23 ( 푛 ) 푛 24 25 26 1 27 퐴푛푛푢푖푡푦 = 1 ― (4) (1 + 푅)푛 28 29 푅 30 31 32 푃푉 = 푐푎푠ℎ 푓푙표푤 ∙ 푎푛푛푢푖푡푦 (5) 33 34 35 푁푃푉 = 푃푉 ― 퐶퐴푃퐸푋 (6) 36 37 38 Where n is the time of the project development, and r is the discount rate or interest rate. 39 40 The equations formulated replace calculating cash flows for every year of the project and 41 42 then adding them up to calculate the PV. 43 44 45 2.2. Real Options Analysis (ROA) 46 47 A ROA or real options valuation (ROV) is performed alongside with the DCF method; ROA 48 49 is applied using PV of free cash flows and NPV of a project calculated with DCF method as 50 51 a basis to estimate different outcomes (Kodukula & Papudesu, 2006). Traditional options 52 53 used in ROA are the option to defer or the option to wait to invest in a project, expand a 54 55 project, and an option to choose between projects (Kodukula & Papudesu, 2006). 56 57 58 11 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 12 of 49

1 2 3 In this study, a binomial tree is used, which is a decision tree. The binomial model can be 4 5 solved using risk-neutral probabilities or market-replicating portfolios, both have the same 6 7 theoretical framework leading to the same solutions but with different mathematical 8 9 approaches (Kodukula & Papudesu, 2006). The inputs required to build a binomial tree and 10 11 12 calculate the option value are the risk-free rate (r), the value of the underlying asset (S0) 13 14 which is the PV of the expected free cash flows based on the DFC method, the cost of 15 16 exercising the option (X) which is the investment required, the life of the option (T), the 17 18 volatility factor (σ) which is a measure of the variability of the underlying asset during its 19 For Peer Review 20 lifetime, and the time frame chosen for the calculations (δt) (Kodukula & Papudesu, 2006). 21 22 In Figure 1, a generic binomial tree of 2 steps is shown. 23 24 25 The up (u) and down (d) factors are functions of the volatility of the underlying asset, and 26 27 they are described as follows: 28 29 30 푢 = exp (휎 훿푡 ) (7) 31 32 33 1 34 푑 = (8) 35 푢 36 37 38 The risk-neutral probability is a mathematical intermediate that allows for discounting the 39 40 cash flows using a risk-free interest rate. The risk-neutral probability (p) is defined as follows: 41 42 43 exp (푟훿푡) ― 푑 푝 = (9) 44 푢 ― 푑 45 46 47 48 49 50 51 52 53 54 55 56 57 58 12 59 60 http://mc.manuscriptcentral.com/sd Page 13 of 49 Sustainable Development

1 2 3 4 5 6 7 8 9 10 11 Figure 1: Generic recombining binomial tree (source: adapted from Kodukula and 12 13 14 Papudesu, 2006). 15 16 17 The option to wait also called the option to defer should be included in every project. A 18 19 company may prefer toFor wait before Peer it invests Review in a project with either negative or marginal 20 21 NPV when it is highly uncertain that the project will achieve a high NPV in the future 22 23 (Kodukula & Papudesu, 2006). An example is a delay in mining a deposit until the market 24 25 conditions are favorable. 26 27 28 3. Case study 29 30 Chile has a large mining industry (Cisternas & Gálvez, 2014; Jane, 2003); mining represents 31 32 9.8% of the gross domestic product, and copper mining is responsible for 8.9% (COCHILCO, 33 34 2019). Chile has a rich territory endowed with porphyric deposits in terms of copper and 35 36 molybdenum (Oyarzún & Oyarzún, 2011). In 2018, Chile produced 5,872 million of fine 37 38 copper tons equivalent to 28.3% of global production, which made it the No. 1 world producer 39 40 41 of copper (SERNAGEOMIN, 2019a). The country is also the second world producer of 42 43 molybdenum, a copper by-product, with 60,248 tons representing 20.4% of the world 44 45 production. 46 47 The production of copper in Chile is concentrated mainly in the northern and central parts of 48 49 50 the country, where 40% of the worldwide reserves of copper are found (SERNAGEOMIN, 51 52 2019a). Copper can be found in sulfide ores and oxides. Sulfide ores are the primary source 53 54 of copper. Copper is obtained from sulfide ores by flotation to concentrate copper; mine 55 56 tailings are the waste of flotation. 57 58 13 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 14 of 49

1 2 3 In Chile, there are 740 tailing storage facilities. Most of them come from copper mining, and 4 5 all are registered in a national cadaster kept by The National Service of Geology and Mining 6 7 - SERNAGEOMIN, out of which geochemical characteristics of 634 tailing deposits are 8 9 available online, 56 chemical elements have been analyzed, including 25 CMs. The content 10 11 12 of silicon dioxide is also analyzed because silicon metal is a CM that can be obtained from 13 14 silicon dioxide (SERNAGEOMIN, 2019c). 15 16 The Antofagasta Region is located in the Atacama Desert in northern Chile, and it holds the 17 18 largest tailings deposits of the country, including the biggest one with the capacity of 4,500 19 For Peer Review 20 21 million tons (SERNAGEOMIN, 2019b). This study is focused on active tailing deposits, i.e., 22 23 those currently used until they reach the legally allowed capacity. These tailing deposits are 24 25 Laguna Seca, Talabre, Esperanza, Sierra Gorda, and Mantos Blancos; the location of these 26 27 mine tailings deposits is shown in Figure 2. The capacity of the active mine tailings analyzed 28 29 in this study is presented in Error! Reference source not found.. 30 31 32 33 34 35 36 37 38 39 40 Figure 2: Active TSF located in the Antofagasta Region, source: (SERNAGEOMIN, 41 42 2019b). 43 44 45 46 47 48 4. Results 49 50 Despite low concentrations of CMs found in mine tailings, these appear as possible sources 51 52 of CMs due to their availability in larger volumes and the ease with which they can be treated 53 54 on-site using the geochemical analysis provided by SERNAGEOMIN, which shows the 55 56 57 58 14 59 60 http://mc.manuscriptcentral.com/sd Page 15 of 49 Sustainable Development

1 2 3 average concentration of 56 chemical elements present in tailing facilities 4 5 (SERNAGEOMIN, 2019c). The total mass of each CM contained in the TSF was calculated 6 7 using its average concentration and the mass of the tailing facility. The in-situ values were 8 9 calculated using current prices, in some cases when it was not possible to find the price of 10 11 12 the CM as metal, the price of the oxide was used, so these are estimates based on the 13 14 available data provided by the United States Geological Survey in their report about 15 16 commodities released in 2019 (U.S. Geological Survey, 2019), the prices that were not 17 18 available in this report were taken from websites providing metal prices (Metal Prices, 2019; 19 For Peer Review 20 Statista, 2019). The average concentration, total mass, and price of each CM of Esperanza 21 22 TSF appear in Error! Reference source not found. as an example. The same calculations 23 24 for the other four TSF analyzed can be found in Supplementary Material. 25 26 27 The profitability of using mine tailings as a source of CMs is analyzed using a ROA to 28 29 produce a CM using the examined TSF. The ROA is illustrated using the option to wait if the 30 31 investment is not an attractive option according to the results of the DFC method. Reasons 32 33 to wait include situations when capital cost and/or operating costs are too high, and the 34 35 prices of CMs are not high enough to obtain a positive NPV for the project. 36 37 38 The example used to illustrate this analysis is the production of scandium metal. Scandium 39 40 was chosen as a CM to be extracted from mine tailings out of Table 2 due to its elevated 41 42 price and the quantity present in the TSF, as well as the availability of data about CAPEX 43 44 and OPEX for a similar project of scandium production. Scandium is not reported to be 45 46 47 recovered from mine tailings; it is produced mainly from primary resources and the principal 48 49 source for scandium is metal imports from China (U.S. Geological Survey, 2019), from REE 50 51 and iron ore processing in Bayan Obo, China (European Commission, 52 53 2017b).Hydrometallurgy processes have been suggested to recover scandium from 54 55 secondary sources. 56 57 58 15 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 16 of 49

1 2 3 4.1. Discounted cash flow method 4 5 Data from an actual scandium feasibility study were used to estimate PV and NPV of a 6 7 mining project consisting of producing scandium (Investorintel, 2014). The following 8 9 considerations were made to adapt the data to the case of mine tailings: concentration of 10 11 12 scandium in the tailing deposits is 26 ppm, ten times lower than in primary sources of 13 14 scandium, mining costs are discounted considering that there is no need to incur these costs 15 16 when dealing with mine tailings, lower production is considered according to the demand for 17 18 scandium metal, an updated price is considered. The input parameters are listed in Error! 19 For Peer Review 20 Reference source not found.. 21 22 23 The DCF method using a risk-adjusted discount rate for five years showed that the PV for 24 25 the project is 195 million USD, an NPV -138 million, for an investment of 333 million USD. If 26 27 the DCF method is screened for a longer period, e.g., ten years, it shows a higher value of 28 29 PV and NPV, 244 million USD, and -89 million USD, respectively. It is essential to point out 30 31 32 that traditional mining projects, planned for 15 or 20 years, are used for calculating NPV 33 34 because they are long term investment projects, but in the case of a project based on mine 35 36 tailings and CMs it is more reasonable to use a shorter period. 37 38 39 4.2. Uncertainty and sensitivity analysis of NPV 40 41 42 To study the effect of uncertainties of input variables and their distribution on NPV results 43 44 and their sensitivity. Sensitivity and uncertainty analysis was performed using Monte Carlo 45 46 simulation in RStudio. NPV was modeled using equations (6) to (9), which is a function of 47 48 the inputs used to calculate the NPV; in this analysis, taxes, and depreciation were included. 49 50 The inputs ranged between a minimum and maximum value, considering a range ± 20%, 51 52 the following values for each output were considered: price 4.4– 6.6 million USD/ton, CAPEX 53 54 268 – 402million USD/ton, OPEX 4400 – 6600 USD/ton, discount rate 0.08 – 0.12, 55 56 57 58 16 59 60 http://mc.manuscriptcentral.com/sd Page 17 of 49 Sustainable Development

1 2 3 production 6.4 – 9.6 ton, and the number of years of the project was constant since NPV is 4 5 highly sensitive to changes in the period of investment. The analysis was made for an 6 7 investment of five and ten years. Fifty thousand simulations were run. 8 9 10 Summary of results shows, for an investment of 5 years, a minimum value of -229.03 and 11 12 13 maximum value of -32.5 for NPV and a mean and median that are almost equal, - 137.33 14 15 and -137.47, respectively. 16 17 18 Histogram and boxplot are graphical methods to test normality. The histogram purpose is to 19 For Peer Review 20 graphically summarize the distribution of a data set; it shows the frequency of NPV results. 21 22 The histogram is presented as supplementary material. 23 24 25 The Boxplot represents the summary and the representation of primary data; it allows to 26 27 visualize the minimum, lower quartile, median, upper quartile, and a maximum of a data set 28 29 (Spitzer, Wildenhain, Rappsilber, & Tyers, 2014). Boxplots as a visualization method enable 30 31 to see characteristics that might not be seen otherwise, and they are a straightforward and 32 33 34 informative method of data interpretation (Sun & Genton, 2011). 3 presents the boxplot, BC 35 36 stands for the base case which is the case using a ± 20% variation for each input, then in 37 38 the rest of the boxplots uncertainty of one input was expanded leaving the other inputs in 39 40 the ±20%previously settled, so in C1 the price was wider, between 3.3 and 7.7 million 41 42 USD/ton, leaving the rest of the variables in the previous range of ±20%. In C2 CAPEX was 43 44 varied between 201 and 469 million USD; in C3 OPEX was varied between 3300 and 7700 45 46 USD/ton; in C4 the return rate was varied between 0.06 and 0.14; and in C5 annual 47 48 production was varied between 4.8 and 11.2 tons. 49 50 51 . 52 53 54 55 56 57 58 17 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 18 of 49

1 2 3 4 5 6 Figure 3: Boxplot of net present value uncertainties for different scenarios of a 5-year 7 8 investment scandium project. 9 10 11 To perform global sensitivity analysis, the function of R-studio Soboljansen implements the 12 13 Monte Carlo estimation of Sobol´-Jansen indices with independent inputs. Figure 4 shows 14 15 the main and total effects of the inputs. It can be seen that the main effect and the total effect 16 17 of each variable are very close which indicates that there are no significant interactions 18 19 For Peer Review 20 between them, and it confirms results obtained by the boxplot that CAPEX in this investment 21 22 case has the most significant effect on NPV. 23 24 25 26 27 28 Figure 4: Sobol´-Jansen indexes for 5- years investment scandium project. 29 30 4.3. ROA using a binomial tree method 31 32 ROA is performed using a binomial decision tree methodology presented by Kodukula and 33 34 35 Papudesu (2006), for an investment of five years and an investment of ten years. The 36 37 input parameters and option parameters are shown in Error! Reference source not 38 39 found.. 40 41 42 For the two cases, five years and ten years, option parameters are the same. Next, the 43 44 development of a decision tree for an investment of five years is explained; the procedure 45 46 is the same for both cases. With the option parameters, the binomial tree is built by 47 48 calculating asset values and options values over the life cycle of the option; asset values 49 50 are obtained after multiplying S0 to u and d raised to the power indicated in each node. Those 51 52 are the numbers on top of every node of the binomial tree. For example, in node S u5, the 53 0 54 expected asset value is 874 million USD. 55 56 57 58 18 59 60 http://mc.manuscriptcentral.com/sd Page 19 of 49 Sustainable Development

1 2 3 Option values are the bottom numbers in the tree. The option to wait expires at the end of 4 5 the binomial tree, so a decision cannot be made after the time that the decision tree takes. 6 7 Option values at year five are calculated as the expected asset value of each node minus 8 9 the cost of exercising the option, which corresponds to the investment. So, for example, in 10

11 5 12 node S0u , the expected asset value is 874 million USD, the investment is 333 million USD, 13 5 14 then the net asset value is 541 million USD. The decision in node S0u would be to invest. 15 16 17 If the option value is negative, then the option value is equal to zero, because a real option 18 19 is a choice, not an obligation.For The Peer option to investReview is exercised at nodes where the option 20 21 value is not zero. 22 23 24 Option values are the numbers below each node, and they are calculated with backward 25 26 induction. Figure 5 and Figure 6 show the binomial tree for five years and ten years’ 27 28 investment, respectively. Each node represents value maximization to invest in that point or 29 30 to wait until the next period; at every node, there is an option to either invest in the project 31 32 33 or the option to wait until the next period, until the option expires, the net asset value, in this 34 5 35 case, represents the NPV. Since the option is evaluated at five-year intervals, in node S0u , 36 37 see Figure 5 the decision is to invest in this point, and the option value is 541 million USD, 38 3 2 39 but in year five, there are also nodes where the option value is zero. In node S0u d , the 40 41 expected asset value is 263 million, for an investment of 333 million, the net asset value is 42 43 -70 million. Hence, the option value in this point is zero, so the decision in this node would 44 45 be not to invest. 46 47 48 49 50 51 Figure 5: Binomial tree for the option to wait for five years to invest in the scandium 52 53 project. 54 55 56 57 58 19 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 20 of 49

1 2 3 Next on the intermediate nodes, the expected asset value for keeping the option open is the 4 5 discounted weighted average of potential future option values using the risk-neutral 6 7 probability that value, e.g., at node S u4 is: 8 0 9 10 푝(푆 푢5) + (1 ― 푝)(푆 푢4푑) ∙ 푒푥푝( ―푟훿푡) (10) 11 0 0 12 13 14 In this node, if the option is exercised the payoff would be 647 million, resulting in a net asset 15 16 value of 314 million, by investing 333 million, since the net asset value obtained by keeping 17 18 the option open is higher, 330 million, then the option to invest is not exercised. In some 19 For Peer Review 20 intermediate nodes, the expected asset value is lower than the investment, which results in 21 22 a net loss, then the decision at that node is to wait, which means that the option value is $0. 23 24 The option valuation binomial tree is completed until year 0. 25 26 27 In each node, the upper numbers represent the expected future values of the underlying 28 29 asset during the option life cycle as it evolves according to its cone of uncertainty, see Figure 30 31 32 5. For example, in year 2, the project is estimated to produce a total payoff between 107 33 34 and 355 million USD, at the end of year five, the payoff is between 44 and 874 million. The 35 36 bottom numbers represent option values on the maximization of investing in that point or 37 38 applying the option to wait until the next period. The option expires in the fifth year because 39 40 it is framed in such a way that it considers competitive forces and uncertainty in the market. 41 42 43 Based strictly on DCF results, the payoff of the project will be 195 million USD, resulting in 44 45 a negative NPV of -38 million USD. If the decision is based exclusively on the NPV result, 46 47 then the decision would be not to invest. 48 49 50 Real options analysis provides an additional value of 33 million USD, considering a net 51 52 53 present value of -138 million USD, the added value with real options analysis is 171 million 54 55 USD. The same analysis was made for a 10-year investment as shown in Figure if the same 56 57 58 20 59 60 http://mc.manuscriptcentral.com/sd Page 21 of 49 Sustainable Development

1 2 3 investment is screened to 10 years, NPV based on the DCF method is -89 million, ROA 4 5 analysis gives an added value of 105 million. 6 7 8 9 10 11 Figure 6: Binomial tree for the option to wait for ten years to invest in the scandium project. 12 13 14 5. Discussion 15 16 Waste valorization is a crucial component that the mining industry must do to shift from a 17 18 linear economy to a circular economy (Khaldoun, Ouadif, Baba, & Bahi, 2016; Kinnunen & 19 For Peer Review 20 Kaksonen, 2019). There are different approaches to mine tailings valorization, e.g., the 21 22 simplest and traditional one is to recover water or to decrease water usage in the tailing 23 24 stage. Other strategies consist in recovering metals from mine tailings or using mine tailings 25 26 27 as a construction material (Ahmari & Zhang, 2012; Lam, Zetola, Ramírez, Montofré, & 28 29 Pereira, 2020). 30 31 The cyclic nature of the supply and demand for specific metals whose production is 32 33 34 extremely concentrated in geographic terms and controlled, and where society cannot solely 35 36 rely on ore mining (Kirchherr, Reike, & Hekkert, 2017) makes sticking to circular economy 37 38 principles, such as effective recycling and processing of secondary sources, a must (Tunsu 39 40 et al., 2019). Mine tailings are now seen as a potential source of several metals and minerals 41 42 (Binnemans et al., 2015; Tunsu et al., 2019). The valorization of mine tailings is at an early 43 44 stage of research; there is a lack of technology that would enable transferring knowledge 45 46 from the laboratory scale to an industrial level (Kinnunen & Kaksonen, 2019). 47 48 49 Results show that tailing storage facilities of copper mines in Chile contain significant 50 51 quantities of CMs. Therefore, the early evaluation of the project's profitability and feasibility 52 53 to recover CMs from mine tailings help in finding alternatives, considering the flexibility given 54 55 by real options, even to postpone the investment. Additionally, it represents a secondary 56 57 58 21 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 22 of 49

1 2 3 source of CMs, decreasing the dependence on reserve extraction (El Wali, Golroudbary, & 4 5 Kraslawski, 2019). 6 7 8 Economic gains from the recovery of CMs come with added benefits of reducing the volume 9 10 of mine waste. Additional benefits can be achieved if, as a result of the processing of mine 11 12 tailings, these can be chemically and physically stabilized, reducing the costs of mine 13 14 closure by decreasing environmental impact and health consequences. A significant 15 16 advantage is to recover water for mining reuse from mine tailings while reducing its reliance 17 18 on seawater (Cisternas & Gálvez, 2018; Ramírez, Cisternas, & Kraslawski, 2017). Another 19 For Peer Review 20 21 essential factor is decreasing energy consumption and greenhouse gas emissions involved 22 23 in pumping seawater to high altitudes for hyperarid locations, where mining companies are 24 25 located (Araya, Lucay, Cisternas, & Gálvez, 2018; Ramírez, Kraslawski, & Cisternas, 2019). 26 27 It is also essential to notice that obtaining CMs from mine tailings reduces the processing of 28 29 comminution stages (Falagán et al., 2017), minimizing energy consumption and greenhouse 30 31 gas emissions. Notably, greenhouse gas emissions can vary for every CM, e.g., in the case 32 33 of phosphorus the greenhouse gas emissions rise gradually in the mining phase and 34 35 increase exponentially in the recycling stage (Rahimpour Golroudbary, El Wali, & 36 37 38 Kraslawski, 2019); in the case of niobium, mining represents 21% of greenhouse gas 39 40 emissions, 72% are generated in the production stage and recycling from scrap accounts 41 42 for only 7% (Rahimpour Golroudbary, Krekhovetckii, El Wali, & Kraslawski, 2019); on the 43 44 other hand, recycling of lithium from lithium-ion batteries generates greenhouse gas 45 46 emissions by 16-20% higher than its primary production (Rahimpour Golroudbary, Calisaya- 47 48 Azpilcueta, & Kraslawski, 2019). 49 50 51 The analysis made on the in-situ value of TSFs of active tailings of northern Chile shows 52 53 that they contain the considerable quantities of several critical metals such as REEs, 54 55 vanadium, cobalt, silicon metal and scandium. The advantages of mine tailings re- 56 57 58 22 59 60 http://mc.manuscriptcentral.com/sd Page 23 of 49 Sustainable Development

1 2 3 processing include the recovery of desirable metals in one location, ensuring land 4 5 reclamation, reduction of landfill areas, diminishing the concentration of harmful compounds, 6 7 and no need to open new mines (Binnemans et al., 2013; Farjana, Huda, Parvez Mahmud, 8 9 & Saidur, 2019; Ganguli & Cook, 2018). 10 11 12 Since mine tailings have been already part-processed, by comminution, the costs, both 13 14 CAPEX and OPEX, of extracting metals from tailings could be attractive in comparison to 15 16 development of a new project based on primary ores (Falagán et al., 2017). For the data 17 18 used, the results of the sensitivity and uncertainty analysis shows that CAPEX has a greater 19 For Peer Review 20 21 influence than the other studied variables. Therefore, it is justified to develop processes and 22 23 technologies that enable lower capital expenses. If time is a variable in an investment 24 25 project, then time and CAPEX are both variables with greater influence. 26 27 28 ROA brings in additional value, so decision-makers can explore alternatives while waiting to 29 30 invest or decide to abandon a project. While waiting for the uncertainty to clear off, to re- 31 32 estimate the project payoff (Arnold, 2014; Kodukula & Papudesu, 2006). If the payoff 33 34 continues to be unfavorable, the decision may be to keep waiting; otherwise, if the conditions 35 36 to invest in scandium production from mine tailings are favorable with a high expected 37 38 39 payoff, the decision would be to invest. ROA calculations are a supplement to traditional 40 41 valuation methods such as DCF based NPV (Brandão & Dyer, 2005; Kodukula & Papudesu, 42 43 2006). 44 45 46 ROA is usually framed in shorter periods than the ones used for long term investments 47 48 (Kodukula & Papudesu, 2006), in this study five and ten-year time horizons were used to 49 50 frame the project; for both periods, NPV is negative, but if the uncertainty clears out, the 51 52 project may be feasible. Framing the investment in a longer period, such as ten years, but 53 54 55 56 57 58 23 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 24 of 49

1 2 3 not as long as twenty years, allows decision-makers to wait to invest in a project when 4 5 investment can be regained. 6 7 8 The methodology presented in this study was validated using copper tailings of northern 9 10 Chile as case study. The geochemical content of tailings may vary from one geographical 11 12 13 area to another depending on the type of mineral deposits. In consequence, the content of 14 15 the valuable elements will be different depending on the mine tailing analyzed. The 16 17 geochemical results of mine tailings analysis are essential to perform profitability analysis of 18 19 their re-processing. For Peer Review 20 21 22 On the limitations of the study, it considers only the economic aspect of valorization of the 23 24 project to obtain CMs from mine tailings. However, it should be mentioned that the 25 26 valorization of the tailing may be profitable in the most cases if environmental, social and 27 28 safety factors are considered, e.g. chemical and physical stabilization, land reclamation, 29 30 social license to operate. As was previously stated by Kinnunen and Kaksonen (2019) 31 32 technology development is still much needed to add value to mine tailings. Therefore, a 33 34 multidisciplinary approach is imperative to develop projects based on re-processing mine 35 36 tailings. Future research should include the assessment of environmental impact of the re- 37 38 39 processing of mine tailings, as they can contain heavy metals that could be scattered during 40 41 the process. Some tailings deposits in Chile need urgent intervention due to the presence 42 43 of one or more heavy metal (Lam, Montofré, et al., 2020). 44 45 The main contribution of this study is to provide a methodology to assess the feasibility of 46 47 48 re-processing of mine tailings by taking into account the flexibility needed for investing in the 49 50 mining projects. This approach is based on use of DCF method combined with ROA and 51 52 sensitivity and uncertainty analysis. 53 54 55 56 57 58 24 59 60 http://mc.manuscriptcentral.com/sd Page 25 of 49 Sustainable Development

1 2 3 6. Conclusions 4 5 Mine tailings need to be adequately stored; otherwise, they can lead to massive 6 7 catastrophes that can affect human settlements and the ecosystem inflicting irreparable 8 9 damage. Proper mine tailings management includes their chemical and physical 10 11 12 stabilization; however, resource recovery can also add profitability to the last stage of mine 13 14 closure. Moreover, the re-processing of wastes contributes to limiting the number of new 15 16 mining projects based on virgin resources, which usually have a considerable environmental 17 18 impact. The assessment of the economic potential of wastes is a crucial activity facilitating 19 For Peer Review 20 the implementation of the circular economy principles. This work presents universal, 21 22 quantitative methodology for the assessment of the economic potential for the recovery of 23 24 valuable elements from mine tailings. 25 26 27 In this study a novel approach to assess the profitability of mine tailings to obtain CM is 28 29 introduced. A framework that includes ROA, sensitivity and uncertainty analysis is proposed, 30 31 which allows for adding flexibility, contrary to traditional valuations tools. 32 33 34 The mine tailings deposits have usually lower concentrations of CMs than ore mines. 35 36 However, the mine tailings are an attractive alternative for exploitation as they are stored in 37 38 deposits with large volumes and hence, having a considerable potential for the production 39 40 of CMs. 41 42 43 The additional factor is the fact that tailings have been already pre-processed, e.g. in 44 45 comminution processes. The case study features Chilean copper mine tailings, to analyze 46 47 the ones with significant volume among the 740 tailing storage facilities present in the 48 49 country. Mine tailings deposits that are currently active were analyzed, using geochemical 50 51 content information provided by The National Service of Geology and Mining 52 53 54 (SERNAGEOMIN). These deposits contain important quantities of rare earths elements, 55 56 cobalt, vanadium, silicon, and barium. Therefore, these copper mine tailings could have 57 58 25 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 26 of 49

1 2 3 substantial economic potential as secondary sources to obtain CMs. It is well-known that 4 5 some CMs are more economically attractive than others, due to their high market price. 6 7 Therefore, demand, supply risk and economic importance of each CM should be considered 8 9 when analyzing the profitability of mine tailings re-processing. 10 11 12 DCF and ROA were applied to analyze the feasibility of producing scandium by re- 13 14 processing the tailing deposits featured in the case study. Additionally, sensitivity and 15 16 uncertainty analysis was performed to study the influence of price of commodity, CAPEX, 17 18 OPEX, and discount rate. These evaluations were used as a complement to the DCF and 19 For Peer Review 20 21 they include Monte Carlo simulation of the NPV and the representation of the results as 22 23 boxplot and Sobol´-Jansen indices. The DCF method showed a negative NPV for 24 25 investments of 5 and 10 years. The results of the sensitivity and uncertainty analysis indicate 26 27 that CAPEX has a greater influence on NPV. ROA adds flexibility and the possibility, as an 28 29 alternative, to choose to wait until the uncertainty is reduced. Based on the calculations, it 30 31 is concluded that using a real options approach to analyze an investment for the production 32 33 of CMs from mine tailings is a useful instrument, considering the high uncertainties in the 34 35 market and the development of process technology. 36 37 38 When traditional methods, such as the DCF method, indicate that a project is not 39 40 economically attractive, a ROA reevaluates the project by considering options such as wait 41 42 until the project is economically feasible. The option of waiting is an interesting alternative 43 44 in the case of resource valorization of mine tailings. It is due to the lack of technological 45 46 47 development and economic business models for treating mine tailings as a source of CMs. 48 49 Therefore, more research on these aspects is still needed. 50 51 Mine tailings as a secondary source for obtaining valuable elements is still a novel option 52 53 54 that needs to be developed on an industrial scale. Therefore, there are many opportunities 55 56 for improvement that may eventually lead to the successful valorization of mining waste. A 57 58 26 59 60 http://mc.manuscriptcentral.com/sd Page 27 of 49 Sustainable Development

1 2 3 multidisciplinary approach is essential to secure the success of re-processing of mine 4 5 tailings to obtain CMs. 6 7 8 References 9 10 Adiansyah, J. S., Rosano, M., Vink, S., & Keir, G. (2015). A framework for a sustainable 11 12 approach to mine tailings management: Disposal strategies. Journal of Cleaner 13 14 Production, 108, 1050–1062. https://doi.org/10.1016/j.jclepro.2015.07.139 15 16 17 Ahmari, S., & Zhang, L. (2012). Production of eco-friendly bricks from copper mine tailings 18 19 through geopolymerization.For Peer Construction Review and Building Materials, 29, 323–331. 20 21 https://doi.org/10.1016/j.conbuildmat.2011.10.048 22 23 24 Alcalde, J., Kelm, U., & Vergara, D. (2018). Historical assessment of metal recovery potential 25 26 from old mine tailings: A study case for porphyry copper tailings, Chile. Minerals 27 28 Engineering, 127(April), 334–338. https://doi.org/10.1016/j.mineng.2018.04.022 29 30 31 Andersson, M., Finne, T. E., Jensen, L. K., & Eggen, O. A. (2018). Geochemistry of a copper 32 33 mine tailings deposit in Repparfjorden, northern Norway. Science of the Total 34 35 Environment, 644, 1219–1231. https://doi.org/10.1016/j.scitotenv.2018.06.385 36 37 38 Araya, N., Kraslawski, A., & Cisternas, L. A. (2020). Towards mine tailings valorization: 39 40 Recovery of critical materials from Chilean mine tailings. Journal of Cleaner Production, 41 42 263(January 2019), 121555. https://doi.org/10.1016/j.jclepro.2020.121555 43 44 45 Araya, N., Lucay, F. A., Cisternas, L. A., & Gálvez, E. D. (2018). Design of Desalinated 46 47 Water Distribution Networks: Complex Topography, Energy Production, and Parallel 48 49 Pipelines. Industrial and Engineering Chemistry Research, 57(30). 50 51 https://doi.org/10.1021/acs.iecr.7b05247 52 53 54 Arnold, T. (2014). A Pragmatic Guide to Real Options. (Palgrave Macmillan, Ed.), A 55 56 57 58 27 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 28 of 49

1 2 3 Pragmatic Guide to Real Options. New York: Palgrave Macmillan. 4 5 https://doi.org/10.1057/9781137391162 6 7 8 Attoh-Okine, N. O., & Ayyub, B. M. (2005). Applied Research in Uncertainty Modeling and 9 10 Analysis. (Springer, Ed.), The International Series of in Intelligent Technologies. 11 12 Boston: Springer. https://doi.org/https://doi.org/10.1007/b101807 13 14 15 Babel, S., Chauhan, R. S., Ali, N., & Yadav, V. (2016). Preparation of phosphate mine 16 17 tailings and low grade rock phosphate enriched bio-fertilizer. Journal of Scientific and 18 19 Industrial ResearchFor. Peer Review 20 21 22 Bhojwani, S., Topolski, K., Mukherjee, R., Sengupta, D., & El-Halwagi, M. M. (2019). 23 24 Technology review and data analysis for cost assessment of water treatment systems. 25 26 Science of the Total Environment. https://doi.org/10.1016/j.scitotenv.2018.09.363 27 28 29 Binnemans, K., Jones, P. T., Blanpain, B., Van Gerven, T., & Pontikes, Y. (2015). Towards 30 31 zero-waste valorisation of rare-earth-containing industrial process residues: A critical 32 33 review. Journal of Cleaner Production, 99, 17–38. 34 35 https://doi.org/10.1016/j.jclepro.2015.02.089 36 37 38 Binnemans, K., Jones, P. T., Blanpain, B., Van Gerven, T., Yang, Y., Walton, A., & Buchert, 39 40 M. (2013). Recycling of rare earths: A critical review. Journal of Cleaner Production, 41 42 51, 1–22. https://doi.org/10.1016/j.jclepro.2012.12.037 43 44 45 Borgonovo, E. (2017). Sensitivity Analysis: An Introduction for the Management Scientist. 46 47 International Series in Operations Research & Management Science. 48 49 https://doi.org/10.1007/978-3-319-52259-3 50 51 52 Borra, C. R., Pontikes, Y., Binnemans, K., & Van Gerven, T. (2015). Leaching of rare earths 53 54 from bauxite residue (red mud). Minerals Engineering, 76, 20–27. 55 56 57 58 28 59 60 http://mc.manuscriptcentral.com/sd Page 29 of 49 Sustainable Development

1 2 3 https://doi.org/10.1016/j.mineng.2015.01.005 4 5 6 Brandão, L. E., & Dyer, J. S. (2005). Decision analysis and real options: A discrete time 7 8 approach to real option valuation. Annals of Operations Research, 135(1), 21–39. 9 10 https://doi.org/10.1007/s10479-005-6233-9 11 12 13 Brandão, L. E., Dyer, J. S., & Hahn, W. J. (2005). Using Binomial Decision Trees to Solve 14 15 Real-Option Valuation Problems. Decision Analysis, 2(2), 69–88. 16 17 https://doi.org/10.1287/deca.1050.0040 18 19 For Peer Review 20 Cacuci, D. G. (2003). Sensitivity and Uncertainty Analysis - Theory (1st Editio). New York: 21 22 Chapman & Hall/CRC. https://doi.org/10.1201/9780203498798 23 24 25 Careddu, N., Dino, G. A., Danielsen, S. W., & Přikryl, R. (2018). Raw materials associated 26 27 with extractive industry: An overview. Resources Policy, 59(September), 1–6. 28 29 https://doi.org/10.1016/j.resourpol.2018.09.014 30 31 32 Ceniceros-Gómez, A. E., Macías-Macías, K. Y., de la Cruz-Moreno, J. E., Gutiérrez-Ruiz, 33 34 M. E., & Martínez-Jardines, L. G. (2018). Characterization of mining tailings in México 35 36 for the possible recovery of strategic elements. Journal of South American Earth 37 38 Sciences, 88(August), 72–79. https://doi.org/10.1016/j.jsames.2018.08.013 39 40 41 CESCO. (2019). Los relaves son una oportunidad para avanzar en una minería de menor 42 43 impacto. Retrieved January 1, 2020, from http://www.cesco.cl/2019/08/16/los-relaves- 44 45 son-una-oportunidad-para-avanzar-en-una-mineria-de-menor-impacto/ 46 47 48 Cisternas, L. A., & Gálvez, E. D. (2014). Chile’s Mining and Chemicals Industries | AIChE. 49 50 Retrieved July 30, 2020, from 51 52 https://www.aiche.org/resources/publications/cep/2014/june/chiles-mining-and- 53 54 chemicals-industries 55 56 57 58 29 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 30 of 49

1 2 3 Cisternas, L. A., & Gálvez, E. D. (2018). The use of seawater in mining. Mineral Processing 4 5 and Extractive Metallurgy Review, 39(1), 18–33. 6 7 https://doi.org/10.1080/08827508.2017.1389729 8 9 10 COCHILCO. (2019). Producto Interno Bruto por Clase de Actividad Económica . Retrieved 11 12 July 26, 2020, from http://www.cochilco.cl:4040/boletin-web/pages/tabla13/buscar.jsf 13 14 15 Collan, M. (2011). Thoughts about selected models for the valuation of real options. Acta 16 17 Universitatis Palackianae Olomucensis. Facultas Rerum Naturalium. Mathematica, 18 19 50(2), 5–12. For Peer Review 20 21 22 Curry, J. A., Ismay, M. J. L., & Jameson, G. J. (2014). Mine operating costs and the potential 23 24 impacts of energy and grinding. Minerals Engineering, 56, 70–80. 25 26 https://doi.org/10.1016/j.mineng.2013.10.020 27 28 29 David, M., & Koch, F. (2019). “Smart is not smart enough!” anticipating critical raw material 30 31 use in smart city concepts: The example of smart grids. Sustainability (Switzerland), 32 33 11(16), 1–11. https://doi.org/10.3390/su11164422 34 35 36 de Koning, A., Kleijn, R., Huppes, G., Sprecher, B., van Engelen, G., & Tukker, A. (2018). 37 38 Metal supply constraints for a low-carbon economy? Resources, Conservation and 39 40 Recycling, 129(November 2017), 202–208. 41 42 https://doi.org/10.1016/j.resconrec.2017.10.040 43 44 45 De Reyck, B., Degraeve, Z., & Vandenborre, R. (2008). Project options valuation with net 46 47 present value and decision tree analysis. European Journal of Operational Research, 48 49 184(1), 341–355. https://doi.org/10.1016/j.ejor.2006.07.047 50 51 52 Dino, G. A., Mehta, N., Rossetti, P., Ajmone-Marsan, F., & De Luca, D. A. (2018). 53 54 Sustainable approach towards extractive waste management: Two case studies from 55 56 57 58 30 59 60 http://mc.manuscriptcentral.com/sd Page 31 of 49 Sustainable Development

1 2 3 Italy. Resources Policy, 59(July 2018), 33–43. 4 5 https://doi.org/10.1016/j.resourpol.2018.07.009 6 7 8 El Wali, M., Golroudbary, S. R., & Kraslawski, A. (2019). Impact of recycling improvement 9 10 on the life cycle of phosphorus. Chinese Journal of Chemical Engineering, 27(5), 1219– 11 12 1229. https://doi.org/10.1016/j.cjche.2018.09.004 13 14 15 European Commission. (2017a). Methodology for establishing the EU list of Critical Raw 16 17 Materials, 25. https://doi.org/10.2873/769526 18 19 For Peer Review 20 European Commission. (2017b). Study on the review of the list of critical raw materials - 21 22 Publications Office of the EU. Retrieved June 10, 2018, from 23 24 https://publications.europa.eu/en/publication-detail/-/publication/08fdab5f-9766-11e7- 25 26 b92d-01aa75ed71a1/language-en 27 28 29 Falagán, C., Grail, B. M., & Johnson, D. B. (2017). New approaches for extracting and 30 31 recovering metals from mine tailings. Minerals Engineering, 106, 71–78. 32 33 https://doi.org/10.1016/j.mineng.2016.10.008 34 35 36 Farjana, S. H., Huda, N., Parvez Mahmud, M. A., & Saidur, R. (2019). A review on the impact 37 38 of mining and mineral processing industries through life cycle assessment. Journal of 39 40 Cleaner Production, 231, 1200–1217. https://doi.org/10.1016/j.jclepro.2019.05.264 41 42 43 Figueiredo, J., Vila, M. C., Matos, K., Martins, D., Futuro, A., Dinis, M. de L., … Fiúza, A. 44 45 (2018). Tailings reprocessing from Cabeço do Pião dam in Central Portugal: A kinetic 46 47 approach of experimental data. Journal of Sustainable Mining, 17(3), 139–144. 48 49 https://doi.org/10.1016/j.jsm.2018.07.001 50 51 52 Ganguli, R., & Cook, D. R. (2018). Rare earths: A review of the landscape. MRS Energy & 53 54 Sustainability, 5, E9. https://doi.org/10.1557/mre.2018.7 55 56 57 58 31 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 32 of 49

1 2 3 Gaspars-Wieloch, H. (2019). Project Net Present Value estimation under uncertainty. 4 5 Central European Journal of Operations Research, 27(1), 179–197. 6 7 https://doi.org/10.1007/s10100-017-0500-0 8 9 10 Gordon, R. B., Bertram, M., & Graedel, T. E. (2007). On the sustainability of metal supplies: 11 12 A response to Tilton and Lagos. Resources Policy, 32(1–2), 24–28. 13 14 https://doi.org/10.1016/j.resourpol.2007.04.002 15 16 17 Guj, P., & Chandra, A. (2019). Comparing different real option valuation approaches as 18 19 applied to a copperFor mine. Peer Resources Review Policy, 61(September 2018), 180–189. 20 21 https://doi.org/10.1016/j.resourpol.2019.01.020 22 23 24 Hällström, L. P. B., Alakangas, L., & Martinsson, O. (2018). Geochemical characterization 25 26 of W, Cu and F skarn tailings at Yxsjöberg, Sweden. Journal of Geochemical 27 28 Exploration, 194(September), 266–279. https://doi.org/10.1016/j.gexplo.2018.09.001 29 30 31 Innocenzi, V., De Michelis, I., Kopacek, B., & Vegliò, F. (2014). Yttrium recovery from 32 33 primary and secondary sources: A review of main hydrometallurgical processes. Waste 34 35 Management, 34(7), 1237–1250. https://doi.org/10.1016/j.wasman.2014.02.010 36 37 38 Insley, M. (2002). A real options approach to the valuation of a forestry investment. Journal 39 40 of Environmental Economics and Management, 44(3), 471–492. 41 42 https://doi.org/10.1006/jeem.2001.1209 43 44 45 Investorintel. (2014). Scandium – the Outlier Rare Earth. Retrieved November 12, 2019, 46 47 from https://investorintel.com/sectors/technology-metals/technology-metals- 48 49 intel/scandium-outlier-rare-earth/ 50 51 52 Jane, N. (2003). Social consequences of mining and present day solutions--Region II in 53 54 Chile highlighted. Sustainable Development, 11(2), 84. Retrieved from http://80- 55 56 57 58 32 59 60 http://mc.manuscriptcentral.com/sd Page 33 of 49 Sustainable Development

1 2 3 proquest.umi.com.ezproxy.mnl.umkc.edu/pqdlink?did=423069461&Fmt=7&clientId=4 4 5 5248&RQT=309&VName=PQD 6 7 8 Jeswiet, J. (2017). Including Towards Sustainable Mining in Evaluating Mining Impacts. 9 10 Procedia CIRP, 62, 494–499. https://doi.org/10.1016/j.procir.2016.06.123 11 12 13 Khaldoun, A., Ouadif, L., Baba, K., & Bahi, L. (2016). Valorization of mining waste and 14 15 tailings through paste backfilling solution, Imiter operation, Morocco. International 16 17 Journal of Mining Science and Technology, 26(3), 511–516. 18 19 https://doi.org/10.1016/j.ijmst.2016.02.021For Peer Review 20 21 22 Khalil, A., Argane, R., Benzaazoua, M., Bouzahzah, H., Taha, Y., & Hakkou, R. (2019). Pb– 23 24 Zn mine tailings reprocessing using centrifugal dense media separation. Minerals 25 26 Engineering, 131(January 2018), 28–37. https://doi.org/10.1016/j.mineng.2018.10.023 27 28 29 Khorasanipour, M., & Jafari, Z. (2017). Environmental geochemistry of rare earth elements 30 31 in Cu-porphyry mine tailings in the semiarid climate conditions of Sarcheshmeh mine 32 33 in southeastern Iran. Chemical Geology, 477(August 2017), 58–72. 34 35 https://doi.org/10.1016/j.chemgeo.2017.12.005 36 37 38 Kinnunen, P. H. M., & Kaksonen, A. H. (2019). Towards circular economy in mining: 39 40 Opportunities and bottlenecks for tailings valorization. Journal of Cleaner Production, 41 42 228, 153–160. https://doi.org/10.1016/j.jclepro.2019.04.171 43 44 45 Kirchherr, J., Reike, D., & Hekkert, M. (2017). Conceptualizing the circular economy: An 46 47 analysis of 114 definitions. Resources, Conservation and Recycling, 127(April), 221– 48 49 232. https://doi.org/10.1016/j.resconrec.2017.09.005 50 51 52 Kodukula, P., & Papudesu, C. (2006). Project Valuation Using Real Options - A practitioner’s 53 54 guide. Fort Lauderdale, Florida: J. Ross Publishing. 55 56 57 58 33 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 34 of 49

1 2 3 Kozlova, M. (2017). Real option valuation in renewable energy literature: Research focus, 4 5 trends and design. Renewable and Sustainable Energy Reviews, 80(June 2016), 180– 6 7 196. https://doi.org/10.1016/j.rser.2017.05.166 8 9 10 Lam, E. J., Montofré, I. M., Álvarez, F. A., Gaete, N. F., Poblete, D. A., & Rojas, R. J. (2020). 11 12 Methodology to prioritize chilean tailings selection, according to their potential risks. 13 14 International Journal of Environmental Research and Public Health, 17(11), 1–16. 15 16 https://doi.org/10.3390/ijerph17113948 17 18 19 Lam, E. J., Zetola, V.,For Ramírez, Peer Y., Montofré, Review Í. L., & Pereira, F. (2020). Making Paving 20 21 Stones from Copper Mine Tailings as Aggregates, 1–14. 22 23 https://doi.org/10.3390/ijerph17072448 24 25 26 Lèbre, É., Corder, G., & Golev, A. (2017). The Role of the Mining Industry in a Circular 27 28 Economy: A Framework for Resource Management at the Mine Site Level. Journal of 29 30 Industrial Ecology, 21(3), 662–672. https://doi.org/10.1111/jiec.12596 31 32 33 Loucks, D. P., & van Beek, E. (2005). Water Resources Systems Planning and Management 34 35 - An Introduction to Methods, Models and Applications. United Nations Educational, 36 37 Scientific and Cultural Organization. 38 39 40 Macías-Macías, K. Y., Ceniceros-Gómez, A. E., Gutiérrez-Ruiz, M. E., González-Chávez, 41 42 J. L., & Martínez-Jardines, L. G. (2019). Extraction and recovery of the strategic 43 44 element gallium from an iron mine tailing. Journal of Environmental Chemical 45 46 Engineering, 7(2), 102964. https://doi.org/10.1016/j.jece.2019.102964 47 48 49 Markovaara-Koivisto, M., Valjus, T., Tarvainen, T., Huotari, T., Lerssi, J., & Eklund, M. 50 51 (2018). Preliminary volume and concentration estimation of the Aijala tailings pond – 52 53 Evaluation of geophysical methods. Resources Policy, 59(May), 7–16. 54 55 56 https://doi.org/10.1016/j.resourpol.2018.08.016 57 58 34 59 60 http://mc.manuscriptcentral.com/sd Page 35 of 49 Sustainable Development

1 2 3 Mathieux, F., Ardente, F., Bobba, S., Nuss, P., Blengini, G., Alves Dias, P., … Solar, S. 4 5 (2017). Critical Raw Materials and the Circular Economy. Background Report. JRC 6 7 Science-for-policy report. (P. office of the E. Union, Ed.) (EUR 28832). Luxembourg. 8 9 https://doi.org/10.2760/378123 10 11 12 Medas, D., Cidu, R., De Giudici, G., & Podda, F. (2013). Geochemical Behaviour of Rare 13 14 Earth Elements in Mining Environments under Non-Acidic Conditions. Procedia Earth 15 16 and Planetary Science, 7, 578–581. https://doi.org/10.1016/j.proeps.2013.03.023 17 18 19 Metal Prices. (2019). For Rare Earth Peer Metals NewsReview | News, research, prices and quotes. 20 21 Retrieved November 10, 2019, from http://mineralprices.com/rare-earth-metals/ 22 23 24 Moran-Palacios, H., Ortega-Fernandez, F., Lopez-Castaño, R., & Alvarez-Cabal, J. V. 25 26 (2019). The Potential of Iron Ore Tailings as Secondary Deposits of Rare Earths. 27 28 Applied Sciences, 9(14), 2913. https://doi.org/10.3390/app9142913 29 30 31 Mudd, G. M. (2010). The Environmental sustainability of mining in Australia: key mega- 32 33 trends and looming constraints. Resources Policy, 35(2), 98–115. 34 35 https://doi.org/10.1016/j.resourpol.2009.12.001 36 37 38 Mun, J. (2002). Real Options Analysis: Tools and Techniques for Valuing Strategic 39 40 Investments and Decisions. Real Options Analysis:Tools and Techniques for valuing 41 42 strategic investments and decisions. https://doi.org/10.1111/j.1574-0862.2006.00138.x 43 44 45 Munier, N. (2014). Risk Management for Engineering Projects. Risk Management for 46 47 Engineering Projects. https://doi.org/10.1007/978-3-319-05251-9 48 49 50 Nelson, R., Howden, M., & Hayman, P. (2013). Placing the power of real options analysis 51 52 into the hands of natural resource managers - Taking the next step. Journal of 53 54 Environmental Management, 124, 128–136. 55 56 57 58 35 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 36 of 49

1 2 3 https://doi.org/10.1016/j.jenvman.2013.03.031 4 5 6 Northey, S. A., Mudd, G. M., Saarivuori, E., Wessman-Jääskeläinen, H., & Haque, N. (2016). 7 8 Water footprinting and mining: Where are the limitations and opportunities? Journal of 9 10 Cleaner Production, 135, 1098–1116. https://doi.org/10.1016/j.jclepro.2016.07.024 11 12 13 Oyarzún, J., & Oyarzún, R. (2011). Sustainable development threats, inter-sector conflicts 14 15 and environmental policy requirements in the arid, mining rich, Northern Chile territory. 16 17 Sustainable Development, 19(4), 263–274. https://doi.org/10.1002/sd.441 18 19 For Peer Review 20 Peelman, S., Kooijman, D., Sietsma, J., & Yang, Y. (2018). Hydrometallurgical Recovery of 21 22 Rare Earth Elements from Mine Tailings and WEEE. Journal of Sustainable Metallurgy, 23 24 4(3), 367–377. https://doi.org/10.1007/s40831-018-0178-0 25 26 27 Peelman, S., Sun, Z. H. ., Sietsma, J., & Yang, Y. (2016). Hydrometallurgical Extraction of 28 29 Rare Earth Elements From Low Grade Mine Tailings. Rare Metal Technology 2016, 30 31 17–29. https://doi.org/10.1007/978-3-319-48135-7_2 32 33 34 Pivorienė, A. (2017). Real Options and Discounted Cash Flow Analysis to Assess Strategic 35 36 Investment Projects. Economics and Business, 30(1), 91–101. 37 38 https://doi.org/10.1515/eb-2017-0008 39 40 41 Rahimpour Golroudbary, S., Calisaya-Azpilcueta, D., & Kraslawski, A. (2019). The Life 42 43 Cycle of Energy Consumption and Greenhouse Gas Emissions from Critical Minerals 44 45 Recycling: Case of Lithium-ion Batteries. Procedia CIRP, 80, 316–321. 46 47 https://doi.org/10.1016/j.procir.2019.01.003 48 49 50 Rahimpour Golroudbary, S., El Wali, M., & Kraslawski, A. (2019). Environmental 51 52 sustainability of phosphorus recycling from wastewater, manure and solid wastes. 53 54 Science of The Total Environment, 672, 515–524. 55 56 57 58 36 59 60 http://mc.manuscriptcentral.com/sd Page 37 of 49 Sustainable Development

1 2 3 https://doi.org/10.1016/j.scitotenv.2019.03.439 4 5 6 Rahimpour Golroudbary, S., Krekhovetckii, N., El Wali, M., & Kraslawski, A. (2019). 7 8 Environmental Sustainability of Niobium Recycling: The Case of the Automotive 9 10 Industry. Recycling, 4(1), 5. https://doi.org/10.3390/recycling4010005 11 12 13 Ramírez, Y., Cisternas, L. A., & Kraslawski, A. (2017). Application of House of Quality in 14 15 assessment of seawater pretreatment technologies. Journal of Cleaner Production, 16 17 148, 223–232. https://doi.org/10.1016/j.jclepro.2017.01.163 18 19 For Peer Review 20 Ramírez, Y., Kraslawski, A., & Cisternas, L. A. (2019). Decision-support framework for the 21 22 environmental assessment of water treatment systems. Journal of Cleaner Production, 23 24 225, 599–609. https://doi.org/10.1016/j.jclepro.2019.03.319 25 26 27 Regan, C. M., Bryan, B. A., Connor, J. D., Meyer, W. S., Ostendorf, B., Zhu, Z., & Bao, C. 28 29 (2015). Real options analysis for land use management: Methods, application, and 30 31 implications for policy. Journal of Environmental Management, 161, 144–152. 32 33 https://doi.org/10.1016/j.jenvman.2015.07.004 34 35 36 Santibañez, C., de la Fuente, L. M., Bustamante, E., Silva, S., León-Lobos, P., & Ginocchio, 37 38 R. (2012). Potential Use of Organic- and Hard-Rock Mine Wastes on Aided 39 40 Phytostabilization of Large-Scale Mine Tailings under Semiarid Mediterranean Climatic 41 42 Conditions: Short-Term Field Study. Applied and Environmental Soil Science, 2012, 1– 43 44 15. https://doi.org/10.1155/2012/895817 45 46 47 Savolainen, J. (2016). Real options in metal mining project valuation: Review of literature. 48 49 Resources Policy, 50, 49–65. https://doi.org/10.1016/j.resourpol.2016.08.007 50 51 52 Savolainen, J., Collan, M., & Luukka, P. (2017). Using a cycle reverting price process in 53 54 modeling metal mining project profitability. Kybernetes, 46(1), 131–141. 55 56 57 58 37 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 38 of 49

1 2 3 https://doi.org/10.1108/K-05-2016-0114 4 5 6 Schatzki, T. (2003). Options, uncertainty and sunk costs: An empirical analysis of land use 7 8 change. Journal of Environmental Economics and Management, 46(1), 86–105. 9 10 https://doi.org/10.1016/S0095-0696(02)00030-X 11 12 13 SERNAGEOMIN. (2017). Geochemistry of Surface Samples of the Tailings Deposits in 14 15 Chile Tailings according to the extracted metal. Retrieved January 1, 2020, from 16 17 https://www.sernageomin.cl/wp-content/uploads/2018/02/Geochemistry-of-Surface- 18 19 Samples-of-the-Tailings-Deposits-in-Chile.pdfFor Peer Review 20 21 22 SERNAGEOMIN. (2019a). Anuario de la Minería de Chile. Retrieved October 31, 2019, from 23 24 https://www.sernageomin.cl/anuario-de-la-mineria-de-chile/ 25 26 27 SERNAGEOMIN. (2019b). Datos Públicos Depósito de Relaves – SERNAGEOMIN. 28 29 Retrieved May 6, 2019, from https://www.sernageomin.cl/datos-publicos-deposito-de- 30 31 relaves/ 32 33 34 SERNAGEOMIN. (2019c). Geoquímica de Superficie de Depósitos de Relaves de Chile 35 36 (01/2019). Retrieved November 5, 2019, from https://www.sernageomin.cl/wp- 37 38 content/uploads/2019/01/24-01-2019Geoquimica-Depositos-Relaves-Chile.pdf 39 40 41 Slade, M. E. (2001). Valuing managerial flexibility: An application of real-option theory to 42 43 mining investments. Journal of Environmental Economics and Management, 41(2), 44 45 193–233. https://doi.org/10.1006/jeem.2000.1139 46 47 48 Spitzer, M., Wildenhain, J., Rappsilber, J., & Tyers, M. (2014). BoxPlotR: A web tool for 49 50 generation of box plots. Nature Methods, 11(2), 121–122. 51 52 https://doi.org/10.1038/nmeth.2811 53 54 55 Statista. (2019). Rare earth elements - Statistics & Facts. Retrieved November 10, 2019, 56 57 58 38 59 60 http://mc.manuscriptcentral.com/sd Page 39 of 49 Sustainable Development

1 2 3 from https://www.statista.com/topics/1744/rare-earth-elements/ 4 5 6 Sun, Y., & Genton, M. G. (2011). Functional boxplots. Journal of Computational and 7 8 Graphical Statistics, 20(2), 316–334. https://doi.org/10.1198/jcgs.2011.09224 9 10 11 Szamałek, K., Konopka, G., Zglinicki, K., & Marciniak-Maliszewska, B. (2013). New potential 12 13 source of rare earth elements. Gospodarka Surowcami Mineralnymi / Mineral 14 15 Resources Management, 29(4), 59–76. https://doi.org/10.2478/gospo-2013-0041 16 17 18 Tayebi-Khorami, M., Edraki, M., Corder, G., & Golev, A. (2019). Re-thinking mining waste 19 For Peer Review 20 through an integrative approach led by circular economy aspirations. Minerals, 9(5). 21 22 https://doi.org/10.3390/min9050286 23 24 25 Tilton, J. E., & Lagos, G. (2007). Assessing the long-run availability of copper. Resources 26 27 Policy, 32(1–2), 19–23. https://doi.org/10.1016/j.resourpol.2007.04.001 28 29 30 Trigeorgis, L. (1993). The Nature of Option Interactions and the Valuation of Investments 31 32 with Multiple Real Options. The Journal of Financial and Quantitative Analysis. 33 34 https://doi.org/10.2307/2331148 35 36 37 Trigeorgis, L. (1996). Real options: Managerial flexibility and strategy in resource allocation. 38 39 Cambridge, Massachusetts: MIT Press. 40 41 42 Tukker, A. (2014). Rare Earth Elements Supply Restrictions: Market Failures, Not Scarcity, 43 44 Hamper Their Current Use in High-Tech Applications. Environmental Science & 45 46 Technology, 48(17), 9973–9974. https://doi.org/10.1021/es503548f 47 48 Tunsu, C., Menard, Y., Eriksen, D. Ø., Ekberg, C., & Petranikova, M. (2019). Recovery of 49 50 51 critical materials from mine tailings: A comparative study of the solvent extraction of 52 53 rare earths using acidic, solvating and mixed extractant systems. Journal of Cleaner 54 55 Production, 218, 425–437. https://doi.org/10.1016/j.jclepro.2019.01.312 56 57 58 39 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 40 of 49

1 2 3 U.S. Geological Survey. (2019). Mineral commodity summaries 2019: U.S. Geological 4 5 Survey. https://doi.org/10.3133/70202434 6 7 8 Wang, C., Harbottle, D., Liu, Q., & Xu, Z. (2014). Current state of fine mineral tailings 9 10 treatment: A critical review on theory and practice. Minerals Engineering, 58, 113–131. 11 12 https://doi.org/10.1016/j.mineng.2014.01.018 13 14 15 Wang, L., Ji, B., Hu, Y., Liu, R., & Sun, W. (2017). A review on in situ phytoremediation of 16 17 mine tailings. Chemosphere. https://doi.org/10.1016/j.chemosphere.2017.06.025 18 19 For Peer Review 20 Wang, T., & Neufville, R. De. (2005). Real Options “ in ” Projects. Engineering. 21 22 https://doi.org/10.1.1.145.4602 23 24 25 Wang, X., Yao, M., Li, J., Zhang, K., Zhu, H., & Zheng, M. (2017). China’s rare earths 26 27 production forecasting and sustainable development policy Implications. Sustainability 28 29 (Switzerland), 9(6). https://doi.org/10.3390/su9061003 30 31 32 Yaksic, A., & Tilton, J. E. (2009). Using the cumulative availability curve to assess the threat 33 34 of mineral depletion: The case of lithium. Resources Policy, 34(4), 185–194. 35 36 https://doi.org/10.1016/j.resourpol.2009.05.002 37 38 39 Zhang, B., Liu, C., Li, C., & Jiang, M. (2014). A novel approach for recovery of rare earths 40 41 and niobium from Bayan Obo tailings. Minerals Engineering, 65, 17–23. 42 43 https://doi.org/10.1016/j.mineng.2014.04.011 44 45 46 Žibret, G., Lemiere, B., Mendez, A. M., Cormio, C., Sinnett, D., Cleall, P., … Carvalho, T. 47 48 (2020). National mineral waste databases as an information source for assessing 49 50 material recovery potential from mine waste, tailings and metallurgical waste. Minerals, 51 52 10(5). https://doi.org/10.3390/min10050446 53 54 55 56 57 58 40 59 60 http://mc.manuscriptcentral.com/sd Page 41 of 49 Sustainable Development

1 2 3 Table 1: Allowed capacity in active tailings storage facilities in Antofagasta Region, 4 5 source:(SERNAGEOMIN, 2019b). 6 7 8 Tailing storage facility (TSF) 106 t (until 23.04.2019) 106 t (allowed capacity) 9 10 Laguna Seca 1,302.24 4,500 11 12 Talabre 1,792.72 2,103.95 13 14 Sierra Gorda 142.80 1,350 15 16 17 Esperanza 240.50 750 18 19 Mantos Blancos For Peer130.71 Review 138.2 20 21 22 23 24 Table 2: Concentration, content, and in-situ values of critical materials in Esperanza tailing 25 26 storage facility. 27 28 Critical materials Average Mass (t) Price CM in-situ value 29 6 30 concentration (USD$/t) (10 USD$) 31 (ppm) 32 Vanadium 160 120,000 30,865 3,704 33 34 Cobalt 11 8,250 72,753 600 35 36 REE – Yttrium 49 36,750 36,000 1,323 37 38 Niobium 21 15,750 21,000 331 39 40 41 Barium 185 138,750 180 25 42 43 REE† – Scandium 26 19,500 5,420,000 105,690 44 45 Hafnium 3.88 2,910 775,000 2,255 46 47 Silicon metal 224,632 168,474,330 3,042 512,563 48 49 REE – Lanthanum 15.96 11,970 2,000 24 50 51 REE – Cerium 32.32 24,240 4,830 117 52 53 REE – Praseodymium 4.31 3,233 92,400 299 54 55 56 57 58 41 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 42 of 49

1 2 3 REE – Neodymium 17.46 13,095 51,000 668 4 5 REE – Samarium 3.65 2,738 14,850 41 6 7 REE – Europium 1.06 795 56,000 45 8 9 REE – Gadolinium 3.43 2,573 22,330 57 10 11 12 REE – Terbium 0.49 368 647,500 238 13 14 REE – Dysprosium 2.87 2,153 280,700 604 15 16 REE – Holmium 0.55 413 38,000 16 17 18 REE – Erbium 1.71 1,283 140,000 180 19 For Peer Review 20 REE – Thulium 0.26 195 not available 0 21 22 REE – Ytterbium 1.6 1,200 not available 0 23 24 REE – Lutetium 0.23 173 1,258,000 217 25 26 REE – mischmetal 85.9 63,630 20,300 1,292 27 28 †RRE: rare earth element 29 30 31 Table 3: Input parameters for the Discounted Cash Flow method 32 33 34 Input parameters Value Unit 35 36 Capital expenses 333 million USD 37 38 Operating expenses 5,700 USD/t 39 40 Price of scandium metal 5.42 million USD/ t 41 42 43 Ore grade 26 ppm 44 45 Production 8 Ton/year 46 47 Discount rate/ interest rate 0.1 48 49 Period 5 &10 years 50 51 Taxes rate 0.35 52 53 54 55 56 57 58 42 59 60 http://mc.manuscriptcentral.com/sd Page 43 of 49 Sustainable Development

1 2 3 Table 4: Input and option parameters for ROA. 4 5 6 7 Input parameters Unit 8

9 S0 (5 years) 195 million USD 10

11 S0 (10 years) 244 million USD 12 13 T (time to expiration) 5/10 years 14 15 X 333 million USD 16 17 σ (volatility) 30 % 18 19 r (risk-freeFor rate) Peer Review5 % 20 21 Δt (time step) 1 22 23 Option parameters 24 25 u (up factor) 1.35 26 27 28 d (down factor) 0.74 29 30 P (risk-neutral probability) 0.51 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 43 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 44 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 27 28 Generic recombining binomial tree (source: adapted from Kodukula and Papudesu, 2006). 29 109x85mm (96 x 96 DPI) 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd Page 45 of 49 Sustainable Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Active TSF located in the Antofagasta Region, source: (SERNAGEOMIN, 2019b). 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 46 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 Boxplot of the net present value uncertainties for different scenarios of a 5-year investment scandium 26 project. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd Page 47 of 49 Sustainable Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 26 Sobol´-Jansen indexes for 5- years investment scandium project. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd Sustainable Development Page 48 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 Binomial tree for the option to wait for five years to invest in the scandium project. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd Page 49 of 49 Sustainable Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Peer Review 20 21 22 23 24 25 Binomial tree for the option to wait for ten years to invest in the scandium project. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/sd