Introduction

The utilization of non-renewable energy sources gives rise to numerous ethical concerns. Human depletion of finite natural resources and the resultant global warming have adverse impacts on the functioning of ecosystems. As these ecosystem functions deteriorate, conventional production and consumption practices continue to steer the world towards an era where the biosphere's viability is uncertain, and its effects on life-sustaining systems are poorly understood. The transition from a fossil fuel-dependent economy to one based on sustainable energy sources has become an indispensable route to ensure the survival of the human species.

At the heart of the renewable energy paradigm, lithium stands as the crucial element for energy storage. Lithium plays a pivotal role as a primary resource for the construction of both existing lithium-ion and potential future generations of batteries.

The world’s largest deposits of lithium lie in brines found underneath salt flats in the desert between Chile, Argentina, and Bolivia (Barandiarán, 2019). It is estimated that the three countries, together referred to as the “Lithium Triangle”, account for almost 75% of the world’s supply beneath their salt flats (Ahmad, 2020). Argentina appears to have advantages not only in terms of its natural conditions but also due to its regulatory environment, especially when compared to its neighboring countries in the region. This favorable regulatory climate makes it attractive for attracting investments focused on exploiting salt flats. Research carried out by the InterAmerican Development Bank concurs that Argentina holds the potential to emerge as the global leader in lithium carbonate production (López et al., 2019).

Unfortunately, although the extraction of lithium can yield significant advantages, it also presents a downside concerning environmental risks and the long-term well-being of local communities. According to Agusdinata et al. (2018), while there is a wealth of research on the exploitation of brine, there has been limited analysis of the actual mining's impact on local communities.

In Argentina, a political and social conflict has emerged pitting proponents of lithium exploitation against those who oppose it. Detractors, including primarily local indigenous communities and environmental experts, argue that the project has not obtained, neither the necessary level of social approval, nor the minimum environmental risk assessments. Meanwhile, supporters, which include certain local communities, policymakers, and companies, view it as an exceptional opportunity that should not be missed.

Although the nature of the conflict has been studied in detail, there has been a lack of focus on a comprehensive solution. This study aimed to address the question of to what extent lithium can be extracted without causing harm to the well-being of individuals or communities while still providing satisfactory economic incentives. We followed the system dynamics methodology, which aims to depict interactive relationships among components within a system and comprehend its evolution over time. Initially, we conducted an experiment with students utilizing a modified version of genetic algorithms, allowing us to unveil the systemic structure of the problem. Subsequently, we proceeded to construct a system dynamics model and conducted numerous simulations to quantify interactions, offering conceptual insights that would otherwise remain undiscovered.

Research background

Lithium

Lithium is an extremely lightweight and malleable metal, known for its high reactivity, which facilitates efficient energy transfer. These characteristics render lithium an ideal choice for crafting compact batteries (Minor Metals Trade Association [MMTA], s.f.). In a technical sense, an ion-lithium battery comprises three essential elements: a cathode, typically made of a metallic oxide like lithium cobalt oxide; an anode typically constructed from graphite; and an electrolyte consisting of lithium salts dissolved within organic solvents (Rodríguez et al., 2020).

Lithium deposits are of three main types: brines and related evaporites, pegmatites, and sedimentary rocks (Gruber et al., 2011). Brines are saline waters with high contents of dissolved salts. They can be classified into continental, geothermal and oil field brines (Pistilli, 2023). Continental is the most common form of lithium-containing brine. Pegmatite deposits are large-grained intrusive igneous rocks originating from the solidification of magma deep within the Earth's crust. Within pegmatites, it is possible to find viable quantities of lithium and other elements (Gruber et al., 2011). Lithium is also present in various sedimentary rock formations, including clay and lacustrine evaporites. Within clay deposits, lithium is a component of clay minerals like smectite, requiring separation through processing methods (Gruber et al., 2011).

According to the U.S. Geological Survey (2023), there are approximately 98 million tons of identified lithium resources¹, whereas lithium reserves² account for 26 million ton (Table 1).

Table 1
World Lithium Resources and Reserves

Source: Own analysis based on U.S. Geological Survey (2023).

From the 35.7 million tons of world’s lithium resources, Bolivia accounts for 31.6 % (21 million). According to Alvarez (2023), despite the potential of the Salar de Uyuni to place Bolivia prominently in the global lithium production landscape, strategic missteps and the complex technical challenges associated with extracting lithium from this unique terrain have hindered the country's ability to establish these resources as proven reserves, thereby assessing the feasibility of their exploitation. As a result, Bolivia is currently absent from the list of lithium reserves.

While the exact number of lithium deposits may vary among experts, it is estimated that both resources and reserves are concentrated in few deposits. According to the findings of Vikström et al.’s research (2013), the ten largest world deposits contain 65% of the known reserves and 70% of world lithium resources.

Social and environmental concerns in Argentina

According to León et al. (2020) the dynamics of lithium exploitation are reigniting discussions and tensions regarding the balance between the national, subnational, and local levels when it comes to sovereignty and the management of natural resources among South American countries. In some cases, certain parties have resorted to violence to safeguard their interests, particularly when they feel marginalized from participating in policymaking processes that directly influence them.

Over the past few years, lithium mining in Argentina has accelerated. While there haven't been any violent incidents to date, there are rising tensions between some local indigenous populations and the local administrations. According to Göbel (2013), locals have mixed feelings about lithium mining initiatives, which are characterized by a blend of indifference, uncertainty, and expectations. Although lithium extraction has an impact on all Indigenous communities, not all of them have the same reactions (Pragier, 2019). In accordance with Pragier’s research, the initiation of lithium extraction projects within local communities has generated two kinds of responses, dependent or autonomous. These perspectives, in turn, have sparked calls for retribution and recognition. These are the cases of Olaroz salt flat (dependent and retribution) and Guayatayoc-Salinas Grandes (autonomous and recognition). Indigenous communities advocating for self-governance and recognition rights primarily emphasize the importance of obtaining Free, Prior, and Informed Consent (FPIC)³ and upholding ancestral rights (Kachi Yupi, 2015). Ancestral rights go beyond being a mere assertion; they are intricately connected to age-old social dynamics. In this sense, Teves and Pasarin (2020) argue that salt in Northwestern Argentina should be seen as a valuable resource for trade and a symbolically significant element in fostering social connections between the Puna and the Quebrada regions. This holds historical and archaeological importance, as it occupies a pivotal role in the dynamic exchange within the Andean trade network.

There are ample justifications for environmental concerns. Lithium mining is established within the salt flats of a high-altitude semi-desert, which constitutes extremely delicate ecosystems (Göbel, 2013). Moreover, high-altitude wetlands' inherent ability to regulate the environment and aid in climate change mitigation and adaptation may be jeopardized by the extraction of lithium (Castillo Díaz, 2023). Hydrologically, salt flats in the Puna region are found within numerous closed basins characterized by internal (endorheic) drainage systems. These salars receive water input from sources such as summer rainfall, surface water runoff, and groundwater inflow.

Water scarcity strongly affect local communities. Communities near mining are most vulnerable to mining expansions, while distant residents experience long-lasting impacts due to lower compensation and delayed groundwater recovery (Liu & Agusdinata, 2021). Water usage in lithium facilities can exceed common consumption by several orders of magnitude. In the case of Salar de Atacama in Chile, water consumption from Lithium-mining was higher by two orders of magnitude compared to any other activity (Liu & Agusdinata, 2020). For the Olaroz salt flat, it is projected that its exploitation will result in a consumption during the production cycle that is five times the total consumption of the local departmental population (Arias Alvarado et al., 2022).

There is a massive number of experts that warn about the lack of understanding of the hydrological dynamics of the salt flats in the face of lithium exploitation (Barandiarán, 2019; Economic Commission for Latin America and the Caribbean [ECLAC], 2023; García et al., 2020; Holzbecher, 2005; Izquierdo & Julieta Carilla, 2017; Liu & Agusdinata, 2020; Liu & Agusdinata, 2021; Marchegiani et al., 2019; Mignaqui, 2019; Sticco et al., 2019; Vera et al., 2023). While new technologies to reduce water consumption and avoid brine evaporation are being developed (Nicolaci et al., 2023; Vera et al., 2023), there are other dangers outside just the freshwater reserves. According to Flexer et al. (2018), the emphasis on the chemistry of brine processing has led to a neglect of the analysis of the overall sustainability of the process. Holzbecher (2005) and Sticco et al. (2019) also consider the risk of eroding the freshwater-saltwater interface and urge to complete the necessary hydrogeological research to identify and map these interfaces.

Collective Intelligence and Genetic Algorithms

Succinctly speaking, collective intelligence (CI) refers to a group of individuals acting together in ways that seem intelligent (Massachusetts Institute of Technology, 2010). According to Por and Atlee (2000), CI is older than humankind itself and manifests its primal forms in the synergies and resilience of ecosystems. CI may take the form of swarm intelligence, like bees (Pratt, 2010), or large numbers of people contributing to a single project (Maher et al., 2010). Yet, in essence, CI is grounded on the concept that large groups of cooperating individuals can produce higher-order intelligence innovations than working separately (Lykourentzou et al., 2011).

Johnson (2000) and Hong and Page (2004) established a strong case for the impact of diversity on CI results. Based on years of experimenting with groups, Johnson confirmed that group's ability responds to an individual ability plus a diversity factor. This factor is a group's property not attributable to individuals. On their side, Hong and Page revealed that groups with high cognitive diversity could outperform expert groups in problem solving. They argue that the different heuristics with which people process information and then make decisions more than compensate their lack of knowledge.

Genetic Algorithms (GAs) are metaheuristic algorithms inspired by biological evolutionary processes, emulating the principle of natural selection and the survival of the fittest (Katoch et al., 2020). Some GA standard applications include strategy planning, robot trajectory planning, Travelling Salesman Problem (TSP) and sequence scheduling, function optimizations, aircraft design and communication networks, machine learning–designing neural networks (Dahiya & Sangwan, 2018).

Standard use of GA starts with a specific population of “chromosomes”, where each chromosome represents a potential solution to the given problem (see figure 1). The algorithm requires a fitness function to be optimized. The algorithm selects the top “fittest” from the population and then crosses over them to obtain offspring. Random mutation further affects the new offspring, who will become the new population. This heuristic is iteratively applied until specific conditions are met. Among various potential conditions, Alam et al. (2020), enumerate several cases to consider: the generated solution meeting minimum criteria, the expected population size being reached, the fulfillment of computational requirements like time and cost, or the discovery of the best fitness solutions.

Figure 1
Basic Structure of a Genetic Algorithm (GA)

Source: extracted from Beg et al. (2011, p. 239).

Collective Intelligence experiments.

Collective intelligence experiments can be classified by adapting the work of Maher et al. (2010) and Lykourentzou et al. (2011) (Figure 2). Maher et al. propose a group intelligence continuum, where higher degrees of collective intelligence correspond to increasing collaboration and synthesis. A synthesized final product of a CI experiment means that it is the result of a synergistic and creative combination of its constituents and is almost impossible to trace individual contributions. Conceptually, if A and B are the input of the participants, an additive (or combined) output would be A+B, whereas a synthesized output would result in AB. Lykourentzou et al. classify CI systems as passive and active, further categorizing the active interventions as competitive, collaborative, and hybrid.

Figure 2
Categorization of collective intelligence experiments.

Source: Own development based on Lykourentzou et al. (2011) and Maher et al. (2010).

Methodology

We designed a mixed methodology using a social experiment and the system dynamics methodology (figure 3).

Figure 3
Methodological framework.

According to Sterman (2000), the primary and crucial phase in modeling is articulating the problem. A model should be purposefully designed to address a specific problem rather than attempting to solve an entire system. The utilization of a genetic algorithm in the social experiment served as a tool to express the problem in systemic terms.

Social Experiment

The experiment aimed to identify opportunities for sustainable lithium exploitation through a collaborative and synthesized effort. An adapted version of genetic algorithms (GA) was designed for the experiment. Thirty high school students divided in five groups took part in the experiment. Before the experiment, they were required to research on about the visions on the subject (See table 2).

Table 2
Genetic Algorithm designed for the experiment with high school students.

Students incorporated both supportive and opposing viewpoints regarding lithium exploitation. Each statement consisted of an assertion along with a rationale. For instance, one might state, “Lithium extraction benefits the country by creating jobs”, while another might argue, “The lithium industry is detrimental to the nation due to soil destruction”. Then they crossed over the rationale, i.e., “creating jobs due to soil destruction”. Then the participants introduced random mutations, such as exchanging positions in the sentence and introducing colors, verbs, and adjectives. For example, “Creating jobs may destroy the soil”. These new statements became the offspring. Participants expanded upon as many statements as possible. Subsequently, they selected only those that made logical sense (fitness function). Next, they combined the statements from each group with those of other groups. This process was repeated until the entire group collectively generated several statements, typically ranging from 10 to 15. (See figure 4)

Figure 4
Heuristic of the proposed Genetic Algorithm to integrate contradictory visions.

Experiment results

While creating the population, the students covered much of the state-of-the-art about lithium exploitation. However, after several loops of crossing over rationales and introducing random mutations, a singular emergent phenomenon arose from this experiment: the selected final statements perfectly described the dynamic structure of the problem, composed by feedback loops, causal relationships, and trade-offs (positive and negative causal relationships at the same time). Conflicting perspectives were replaced by a systemic structure that could be modelled and simulated. The final statements and the causal-loop diagram are shown in figures 5 and 6 respectively.

Figure 5
Set of final statements after several loops of crossing-over

Figure 6
Causal loop diagram of lithium exploitation

System Dynamics Model

Model Subsystems

Given their intrinsic interconnectedness within the model, it is more convenient for explanatory purposes to present the variables in subsystems (See Figure 7)

Figure 7
Subsystem diagram for modeling the lithium sustainable exploitation.

In the figure 7, Arrows denote the existence of causal links between the different variables or components of each group. The model will be transformed into a stock-and-flow diagram and then simulated using Powersim Studio 10 Professional simulation software.

Model Stock & Flow Diagram

The causal loop diagram enhances conceptual comprehension and might be viewed as a qualitative model (Liu et al., 2019). Incorporating quantitative modeling through stock-and-flow models, enriched by simulation capabilities, aids in assessing the effects of various scenarios and lends a sense of realism to the systemic structure of the system.

The suggested lithium extraction model is theoretical, although it draws on reported information from the Olaroz and Cauchari-Olaroz projects (Worley Parsons, 2011; Hydrominex Geoscience, 2022; Burga et al., 2019; Houston & Gunn, 2011). Lithium production is based on brine evaporation ponds.

The time horizon for the simulation is established at 40 years for the base case and extended to 50 years in cases where production cannot meet its maximum potential.

Stock & Flow Diagram of The Mineral Resources and Reserves Subsystem

All geological variables, including sediments reserves, porosity, lithium metal concentration and lithium carbonate reserves are based on geological research by Worley Parsons (2011), Burga et al. (2019) and Hydrominex Geoscience (2022).(See Table 3 and Figure 8)

Table 3
Variables of Mineral Resources and Reserves

Figure 8
Measured and Indicated Resources Stock & Flow Diagram

Stock & Flow Diagram of The Investments and Production Capacity Subsystem (See Table 4 and Figure 9)

Table 4
Variables of Investments and Production Capacity

Figure 9
Investments & Production Capacity Stock & Flow Diagram

Stock & Flow Diagram of The Water Footprint Subsystem

The projected water footprint draws from the research conducted by Smakhtin et al. (2004), Mignaqui (2019), and Arias Alvarado et al. (2022). According to the findings of Arias Alvarado et al., the water footprint for lithium carbonate (Li2CO3) is calculated at 584.1 m³ per ton, with freshwater accounting for 8% of this total, equivalent to 46.73 m³ per ton. Smakhtin et al. (p.9) measure water scarcity by creating the Water Stress Indicator or WSI as follows:

[1]

In the equation, MAR corresponds to Mean Annual Runoff and EWR to Environmental Water Requirements. Smakhtin et al. (2004) proposes a range between 20% to 50% of MAR as EWR, or water requirements for ecosystem services. Mignaqui (2019) recommends a 25% for the Puna, the number selected for the simulation. On the other side, MAR is calculated based on Mignaqui (p.48):

[2]

Annual Rainfall in the Olaroz basin is approximately 50 mm (Burga et al., 2019, p. 289). 5% refers to the percentage of rainfall that transforms into surface runoff. According to Smakhtin et al. (2004), the value of WSI determines the degree of water as can be see in table 5, table 6 and figure 10:

Table 5
Categorization of environmental water scarcity

Source: Extracted from Smakhtin et al. (2004, p. 10).

Table 6
Variables of Water Footprint

Figure 10
Water Footprint Stock & Flow Diagram.

Stock & Flow Diagram of The Social License Subsystem

Social License is quantified as a continuous variable within a range of 0 to 1, where a value of 1 indicates complete acceptance by residents and 0 indicates total rejection (see Table 7 and figure 11) . An intermediate value, for instance, 0.65, signifies an average support level of 65%. A higher WSI reduces social license, while employment enhances social license, especially if it has a multiplying effect on local economy. Direct employment is expected to be around 780 (Secretaría de Minería, s.f.; Minera Exar, 2022), whereas Indirect employment has a multiplier of 4,5 (Minera Exar, 2022). Local population is estimated to be 4000 and the starting Social License is estimated to be 0,4.

Table 7
Variables of Social License

Figure 11
Social License Stock & Flow Diagram.

Stock & Flow Diagram of The Project Economics Subsystem

Lithium exploitation projects have internal rates of return (IRR) that range from 24% to 60%, and payback periods from 1, 5 to 5 years (Ibarra-Gutiérrez et al., 2021). Olaroz expansion project estimates a Lithium Carbonate price of USD 14.000 per tonne (Hydrominex Geoscience, 2022). Lithium prices are very unlikely to drop since project pipeline is insufficient to meet future demand (S&P Global Market Intelligence, 2022). Different studies of Olaroz and Caucharí-Olaroz projects placed estimated IRR somewhere between 30% to 40%, a payback period of 2,5 to 5 years and a NPV @8% between USD 2,5 to 3,0 billion (Burga et al., 2019; Hydrominex Geoscience, 2022). The model recreates these numbers as a validation test (See Table 8 and figure 12)

Table 8
Variables of Project Economics

Figure 12
Project Economics Stock & Flow Diagram

Stock & Flow Diagram of The Royalties, Taxes, and Export Retentions Subsystem

Mining activity in Argentina, including prospecting and exploration, is primarily governed by the National Constitution. Following that, the Mining Code, Article 75, Clause 12, comes into effect. Lastly, the Mining Investment Law (Law No. 24.196/93, amended by Law No. 25.429/01), applies. According to these references, provincial royalties are expected to be around 2,3% (Worley Parsons, 2011, p. xv) and export retentions of 4,5% (Freytes et al., 2022). It is expected both a higher federal and provincial support with higher income (See table 9 and Figure 13)

Table 9
Variables of royalties, taxes, and export retentions

Figure 13
Royalties, taxes, and export retentions Stock & Flow Diagram.

Discussion

Base Case

The base case simulation depicts the expected behavior of the system. In a 40-year simulation, the project extracts and refines almost 1,5 million tons of lithium carbonate. The project yields an IRR of 36,4% and a NPV of 2,1 billion USD. Payback time is 5,21 years. Due to employment opportunities, social license increases from 0,4 to a maximum of 0,64, a 1,73-multiplying effect. (See figure 14)

Figure 14
Main indicators of the Base Case simulation.

Source: System Dynamics Model.

It's worth noting that the current total value of state economic benefits, which includes both federal and provincial contributions, stands at approximately 2.0 billion USD, a figure remarkably close to the project's Net Present Value (NPV). With a salt flat surface o 1,5 Km² and an annual rainfall of 50 mm, the WSI slightly exceeds the threshold value of 0,6, yet remains most of the time in the heavily exploited range. The social license is sustained at a relatively high level, primarily attributed to job opportunities and positive externalities among the community. The benefits of job opportunities outweigh potential challenges related to water stress. The support from the province remains robust. The simulation depicts significant potential water stress with substantial social and federal benefits (See figure 15)

Figure 15
Behavior of key variables of the Base Case simulation with 50 m³/ton water consumption.

Source: System Dynamics Model.

Alternative Scenarios

There is a potential scenario where water consumption is considerably higher, for example, at 80 m³ per ton. In such a situation, the outcomes can vary significantly.(See figure 16, 17 and 18)

Figure 16
Main indicators of the 80 m³/ton water consumption scenario.

Figure 17
Behavior of key variables and water stress simulating with 80 m³/ton water consumption.

Source: System Dynamics Model.

In such a scenario (see figure 17), the project essentially pushes water stress to its extreme, well into overexploitation territory, while the project's economic aspects remain unchanged. Consequently, social acceptance drops because of water scarcity. Under the influence of provincial revenue, the local government moderately reduces its support but still aims to maintain lithium production at the current rate since authorities do not employ water stress as a means to regulate production.

If provincial authorities decide to limit production as WSI surpasses the danger threshold, the system behaves seeking an equilibrium goal (figures 18 and 19).

Figure 18
Main indicators of the 80 m³/ton water consumption scenario and production modulated by provincial authorities due to water stress.

Figure 19
Behavior of key variables of the simulation with 80 m³/ton water consumption and production modulated by provincial authorities due to water stress.

(See figure 19) The model depicts an oscillating behavior (cycles of 12-year period) due to the balancing loop to correct water stress with the presence of significant delays to adjust both expectations and production. Economically, the project repays almost in the same period that the base case. NPV reaches 1,68 billion USD and IRR results in 34,2%. After a 40-year simulation, there are many years left of production since remaining sediments reach 18 %. The water stress diminishes significantly, yet the project remains a significant portion of the simulation in the heavily exploited region. Social license and provincial support follow the oscillating behavior yet indicating that both stakeholders still receive significant benefits.

In all previous scenarios, it is assumed that LiCO4 production remains viable even in conditions of high-water stress. The subsequent scenario illustrates a situation where annual rainfall decreases by 50%, imposing a physical constraint on LiCO4 production (see figure 20).

Figure 20
Key indicators for the scenario with a water consumption of 50 m³/ton and a 50% reduction in annual rainfall, imposing a physical limitation on production.

Figure 21
Behavior of key variables for the scenario with a water consumption of 50 m³/ton and a 50% reduction in annual rainfall, imposing a physical limitation on production.

(See figure 21) In this situation, all stakeholders experience reduced benefits. The NPV stands at 0.95 billion USD. Although the IRR exceeds 22%, more than 40% of sediments are still available for exploitation. Despite provincial and physical constraints, the project maintains a positive NPV, and water stress remains predominantly in heavily exploited region.

Policy Design

There is still the possibility of exploiting lithium without stressing the water availability. It’s possibly to produce at the rate that drives WSI to the 0,6 value. That would be the maximum production rate. Using the equations (1) and (2), it is possible to calculate the maximum production as follows:

[3 ]

(See figure 22 and 23)

Figure 22
Main indicators of the simulation with 50 m³/ton water consumption scenario and a maximum production policy based on a threshold WSI of 0,6.

Figure 23
Behavior of main variables of the simulation with 50 m³/ton water consumption scenario and a maximum production policy based on a threshold WSI of 0,6.

The simulation spanning 40 years indicates the feasibility of lithium carbonate (LiCO4) exploitation without causing undue strain on water availability, maintaining a positive Net Present Value (NPV) of 1.32 billion USD and a positive Internal Rate of Return (IRR) of 23.6%. Although the social impact is slightly lower than the base case, it remains significant. The project still possesses several years of production potential, offering opportunities to enhance profitability. Throughout the project's duration, the Water Stress Index (WSI) consistently remains within the moderately exploited range. Repeating the same process for the 80 m³/ton scenario, NPV reaches 0,67 billion USD and IRR remains positive at 17,5% (see figure 24).

Figure 24
Main indicators of the simulation with 80 m³/ton water consumption scenario and a maximum production policy based on a threshold WSI of 0,6.

In all scenarios with a production cap, both social acceptance and government support remain strong. Table 10 presents a sensitivity analysis by varying WSI thresholds under the 80 m³/ton scenario. In every case, the Net Present Value (NPV) is positive, and water withdrawals stay below the overexploitation range.

Table 10
Sensitivity analysis based on varying WSI thresholds under the 80 m³/ton scenario

Source: System Dynamics model.

Model validation

The model's validation process adhered to the criteria set out by Barlas (1996) and Sterman (2000). Structural assessments, including unit consistency checks, sensitivity analyses, and evaluations under extreme conditions, were conducted, revealing the model's robustness. With more than 230 variables encompassing diverse units such as operational, technical, financial, and motivational, the model demonstrates coherent interactions. It accurately replicates the anticipated operational and financial performance of a real-world case, drawing from comprehensive technical data as a reference. Sensitivity analyses exhibit consistent patterns and unveil unexpected insights. The incorporation of provincial support into the lithium production loop results in oscillatory behavior, while capping production diminishes the project's financial resilience, though the impact remains limited. If lithium price drops significantly and operating profitability falls below zero, production will eventually stop. Social support highlights the tension between employment opportunities and water stress, indicating a degree of tolerance and a clear threshold for rejection. Boundary and extreme condition tests consistently affirm model behavior. Following Sterman (2000) guidelines about coherence of physical laws, if rainfall reaches zero, production becomes unfeasible. When remaining sediments approach to zero, no more production is feasible. In cases of drought, if production continues without interruption, social approval eventually diminishes to zero, paralleled by a decline in provincial support.

Conclusion

According to our simulations, lithium extraction in the Puna region entails uneven risks for the involved parties. While the model indicates consistent project profitability and financial returns across all simulated scenarios, significant water stress is evident in all cases except for those relying on a WSI of 0.6 for production. In the face of production limitations based on water scarcity, the expected IRR reaches 24%, almost the same IRR as the preliminary economic assessment (Burga et al., 2019, p. xvi). The simulation also indicates that achieving both profitability and sustainable lithium exploitation is possible through a gradual production increase, contingent upon considerations of water stress.

Both the federal and provincial governments accumulate substantial economic benefits. In aggregate, the present value of these benefits is nearly equivalent to the net present value (NPV) of the company. Proportionally, the federal government retains just over 80% of all government benefits. Simultaneously, lithium exploitation provides residents with employment opportunities and has a noteworthy positive impact on the local economy. Nevertheless, the model indicates the presence of a short-term tolerance threshold, primarily attributed to complexities related to water stress. It suggests that increased salaries and improved livelihoods may not sufficiently compensate for the depletion of water resources needed to sustain ecosystem services, particularly in cases of overexploitation. Under such circumstances, the advancement of lithium mining can only occur if both the government and companies choose to overlook the concerns of the residents.

The model adopts the concept of water stress due to the absence of specific studies on the hydrological dynamics of the region where the saline is located. With this precise information, it is possible to directly model the physical dynamics of water reserves and the risks associated with their exploitation. Despite the model's validation based on experts’ generally accepted criteria, historical corroboration was not entirely possible. Certain historic information such as project economics and operating data have been successfully recreated. However, other variables could have not been validated due to incomplete or complete lack of information. The combined efforts of researchers and the data gathered during this study reveal that local communities are not a homogenous group. They comprise both local residents and indigenous native communities, some of whom endorse lithium exploitation, while others are opposed to it. Future research should focus on exploring into the intricate dynamics that intertwine these communities.

References

Agusdinata, D. B., Liu, W., Eaking, H. y Romero, H. (2018). Socio-environmental impacts of lithium mineral extraction: towards a research agenda. Environmental Research Letters, 13(12), 123001. https://doi.org/10.1088/1748-9326/aae9b1

Ahmad, S. (15 de enero de 2020). The Lithium Triangle: Where Chile, Argentina, and Bolivia Meet. Harvard international Review. https://hir.harvard.edu/lithium-triangle/

Alam, T., Qamar, S., Dixit, A. y Benaida, M. (2020). Genetic Algorithm: Reviews, Implementations, and Applications. International Journal of Engineering Pedagogy. https://doi.org/10.22541/au.159164762.28487263

Allkem. (2023). Annual Report 2023.

Alvarez, J. P. (13 de julio de 2023). Latin America’s Share of Global Lithium Production Seen Declining. Bloomberg Línea. https://www.bloomberglinea.com/english/latin-americas-share-of-global-lithium-production-seen-declining/#:~:text=Although%20the%20Salar%20de%20Uyunito%20know%20how%20feasible%20their

Arias Alvarado, P. V., Diaz Paz, W. F., Salas Barboza, A. G., Seghezzo, L. y Iribarnegaray, M. A. (2022). Huella hídrica como indicador del consumo de agua en la minería de litio en la Puna argentina. Avances en Energías Renovables y Medio Ambiente, 26, 223-234. https://portalderevistas.unsa.edu.ar/index.php/averma/article/view/3833

Barandiarán, J. (2019). Lithium and development imaginaries in Chile, Argentina and Bolivia. World Development, 113, 381-391. https://doi.org/10.1016/j.worlddev.2018.09.019

Barlas, Y. (1996). Formal aspects of model validity and validation in system dynamics. System Dynamics Review, 12(3), 183-210. https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1727(199623)12:3%3C183::AID-SDR103%3E3.0.CO;2-4

Beg, S., Khan, A., Nauman, U. y Mohsin, S. (2011). Performance Evaluation of Bionomic Algorithm (BA) in Comparison with Genetic Algorithm (GA) for Shortest Path Finding Problem. I International Journal of Computer Science Issues, 8(2), 238-242.

Burga, D., Burga, E., Genck, W. y Weber, D. (2019). Updated mineral resource estimate for the Cauchari-Olaroz Project, Jujuy Province, Argentina (Reporte Técnico NI 43 – 101). Lithium Americas. https://www.sec.gov/Archives/edgar/data/1440972/000156459019010645/lac-ex991_6.htm

Castillo Díaz, M. (2023). Humedales altoandinos y puneños: oro blanco, saqueo verde. En P. Marchegiani y A. M. Nápoli (Comps.), Informe Ambiental 2023. El laberinto de las transiciones: aportes para pensar una salida socioecológica en tiempos de crisis (pp. 197-211). Fundación Ambiente y Recursos Naturales.

Dahiya, A. y Sangwan, S. (2018). Literature Review on Genetic Algorithm. International Journal of Research, 5(16), 1142-1146.

Economic Commission for Latin America and the Caribbean . (2023). Lithium extraction and industrialization: Opportunities and challenges for Latin America and the Caribbean. United Nations.

Flexer, V., Baspineiro, C. y Galli, C. (2018). Lithium recovery from brines: A vital raw material for green energies with a potential environmental impact in its mining and processing. Science of The Total Environment, 639, 1188-1204. https://doi.org/10.1016/j.scitotenv.2018.05.223

Freytes, C., Obaya, M. y Delbuono, V. (2022). Federalismo y desarrollo de capacidades productivas y tecnológicas en torno al litio. Fundar

García, R., Kruse, E., Etcheverry, R., Tessone, M. y Moreira, P. (2020). Características hidrogeológicas de los salares en la Puna Argentina. En F. J. Díaz (Ed.), El Litio en La Argentina: Visiones y Aportes Multidisciplinarios desde la UNLP (pp. 49-59). Universidad Nacional de La Plata. http://sedici.unlp.edu.ar/handle/10915/121331

Göbel, B. (2013). La minería del litio en la Puna de Atacama: interdependencias transregionales y disputas locales. Iberoamericana, 13(49), 135-149. https://doi.org/10.18441/ibam.13.2013.49.135-149

Gruber, P., Medina, P., Keoleian, G., Kesler, S., Everson, M. y Wallington, T. (2011). Global Lithium Availability: A Constraint for Electric Vehicles? Journal of Industrial Ecology, 15(5), 760-775. https://doi.org/10.1111/j.1530-9290.2011.00359.x

Holzbecher, E. (2005). Groundwater flow pattern in the vicinity of a salt lake. Hydrobiología, 532, 233–242. https://doi.org/10.1007/s10750-004-6421-7

Hong, L. y Page, S. (2004). Groups of diverse problem solvers can outperform groups of high-ability problem solvers. Proceedings of the National Academy of Sciences of the United States of America, 101(46), 16385–16389. https://doi.org/10.1073/pnas.0403723101

Houston, J. y Gunn, M. (2011). Technical Report on the Salar de Olaroz Lithium-Potash Project. Jujuy Province, Argentina [Reporte Técnico NI 43-101]. Orocobre Ltd.

Hydrominex Geoscience. (2022). Olaroz Lithium Facility Stage 2 Technical Study. Report for ALLKEM.

Ibarra-Gutiérrez, S., Bouchard, J., Laflamme, M. y Fytas, K. (2021). Project economics of lithium mines in Quebec: A critical review. The Extractive Industries and Society, 8(4), 100984. https://doi.org/10.1016/j.exis.2021.100984

Izquierdo, A. y Julieta Carilla, J. (2017). Elaboración del Informe de Inventario de Humedales de la Cuenca Olaroz – Cauchari en la provincia de Jujuy – Nivel 2. Instituto de Ecología Regional. CONICET.

Johnson, N. (2000). Developmental Insights into Evolving Systems: Roles of Diversity, Non-Selection, Self-Organization, Symbiosis [Ponencia]. Seventh International Conference on Artificial Life. Los Alamos National Laboratory. Oregon, USA.

Kachi Yupi. (2015). Huellas de Sal: Procedimiento de consulta y consentimiento previo, libre e informado para las comunidades indígenas de la Cuenca de Salinas Grandes, y Laguna de Guayatayoc. Comunidad de Quera y Aguas Calientes.

Katoch, S., Chauhan, S. y Kumar, V. (2020). A review on genetic algorithm: past,present, and future. Multimedia Tools and Applications, 80(5), 8091–8126. https://doi.org/10.1007/s11042-020-10139-6

León, M., Muñoz, C. y Sánchez, J. (Eds.). (2020). La gobernanza del litio y el cobre en los países andinos. Comisión Económica para América Latina y el Caribe .

Liu, W. y Agusdinata, D. B. (2020). Interdependencies of lithium mining and communities sustainability in Salar de Atacama, Chile. Journal of Cleaner Production, 260, 120838. https://doi.org/10.1016/j.jclepro.2020.120838

Liu, W. y Agusdinata, D. (2021). Dynamics of local impacts in low-carbon transition: Agent-based modeling of lithium mining-community-aquifer interactions in Salar de Atacama, Chile. The Extractive Industries and Society, 8(3), 100927. https://doi.org/10.1016/j.exis.2021.100927

Liu, D., Li, G., Hu, N. y Ma, Z. (2019). Applications of real options on the decision-making of mining investment projects using the system dynamics method. IEEE Access, 7, 46785-46795. https://doi.org/10.1109/ACCESS.2019.2909128

López, A., Obaya, M., Pascuini, P. y Ramos, A. (2019). Litio en Argentina: Oportunidades y desafíos para el desarrollo de la cadena de valor. Banco Interamericano de Desarrollo. http://dx.doi.org/10.18235/0001553

Lykourentzou, I., Vergados, D., Kapetanios, E. y Loumos, V. (2011). Collective intelligence systems: classification and modeling. Journal of Emerging Technologies in Web Intelligence, 3(3), 217-226. https://doi.org/10.4304/jetwi.3.3.217-226

Maher, M., Paulini, M. y Murty, P. (2010). Scaling Up: From Individual Design to Collaborative Design to Collective Design. En J. S. Gero (Ed.), Design Computing and Cognition '10 (pp. 581-600). Springer.

Marchegiani, P., Höglund, J. y Gómez, L. (2019). Extracción de litio en Argentina: un estudio de caso sobre los impactos sociales y ambientales. Fundación Ambiente y Recursos Naturales (FARN).

Massachusetts Institute of Technology. (23 de septiembre de 2010). Thomas Malone on collective intelligence [Archivo de Vídeo]. YouTube. https://www.youtube.com/watch?v=CbR6RaU5SX0

Mignaqui, V. (2019). Puna, litio y agua: estimaciones preliminares para reflexionar sobre el impacto en el recurso hídrico. Revista de Ciencias Sociales, 10(36), 37-55. http://ridaa.unq.edu.ar/handle/20.500.11807/3508

Minera Exar. (2022). Informe de Sostenibilidad 2022.

Minor Metals Trade Association. (s.f.). Lithium. Recuperado el 29 de octubre de 2023 de https://mmta.co.uk/metals/li/#:~:text=Another%20interesting%20property%20of%20Lithiuman%20efficient%20heat%20transfer%20medium.

Nicolaci, H., Young, P., Snowdon, N., Rai, A., Chen, T., Zhang, J., Lin, Y., Bailey, E., Shi, R. y Zheng, N. (27 de abril de 2023). Direct Lithium Extraction: A potential game changing technology. Goldman Sachs.

Pistilli, M. (13 de julio de 2023). Types of Lithium Brine Deposits. Investing News Network. https://investingnews.com/daily/resource-investing/battery-metals-investing/lithium-investing/lithium-deposit-types-brine-pegmatite-and-sedimentary/

Por, G. y Atlee, T. (2000). Collective intelligence as a field of multi-disciplinary study and practice [Ponencia]. Collective Intelligence Conference.

Pragier, D. (2019). Comunidades indígenas frente a la explotación de litio en sus territorios: contextos similares, respuestas distintas. Polis, 52. http://journals.openedition.org/polis/16838

Pratt, S. C. (2010). Collective Intelligence. En M. Breed y J. Moore (Eds.), Encyclopedia of Animal Behavior (pp. 303-309). Academic Press.

Rodríguez, A., Ortiz, M. y Thomas, J. E. (2020). Baterías de ion litio: presente y futuro. En F. J. Díaz (Ed.), El Litio en La Argentina: Visiones y Aportes Multidisciplinarios desde la UNLP (pp. 181-196). Universidad Nacional de La Plata. http://sedici.unlp.edu.ar/handle/10915/121516

Secretaría de Minería. (s.f.). Empleo Minero. Tableau Public. Recuperado el 1 de junio de 2023 de https://public.tableau.com/app/profile/sec.mineria/viz/EmpleoMinero/EmpleoRubro

Smakhtin, V., Revenga, C. y Döll, P. (2004). Taking into account environmental water requirements in global-scale water resources assessments (Comprehensive Assessment Research Report 2). International Water Management Institute.

S&P Global Market Intelligence. (10 de noviembre de 2022). Lithium project pipeline insufficient to meet looming major deficit. S&P Global. https://www.spglobal.com/marketintelligence/en/news-insights/research/lithium-project-pipeline-insufficient-to-meet-looming-major-deficit

Sterman, J. (2000). Business Dynamics. Systems Thinking and Modelling for a Complex World. McGraw-Hill Higher Education.

Sticco, M., Scravaglieri, P. y Damiani, A. (2019). Estudio de los Recursos Hídricos y el Impacto por la Explotación Minera de Litio: Cuenca Salinas Grandes y Laguna Guayatayoc. Provincia de Jujuy. Fundación Ambiente y Recursos Naturales (FARN).

Teves, L. S. y Pasarin, L. (2020). Comunidades locales, manejo del medio ambiente y recursos minerales. Contextos etnográficos y la exploración del litio en el noroeste argentino (NOA). En F. J. Díaz (Ed.), El Litio en Argentina: Visiones y Aportes Mulitdisciplinarios desde la UNLP (pp. 70-81). Universidad Nacional de La Plata. http://sedici.unlp.edu.ar/handle/10915/121523

U.S. Geological Survey. (2023). Mineral Commodity Summaries 2023.https://doi.org/10.3133/mcs2023

Vera, M. L., Torres, W. R., Galli, C. I., Chagnes, A. y Flexer, V. (2023). Environmental impact of direct lithium extraction from brines. Nature Reviews Earth & Environment, 4(3), 149-165. https://doi.org/10.1038/s43017-022-00387-5

Vikström, H., Davidsson, S. y Höök, M. (2013). Lithium availability and future production outlooks. Applied Energy, 110, 252-266. https://doi.org/10.1016/j.apenergy.2013.04.005

Worley Parsons. (30 de abril de 2011). NI 43 - 101 Technical Report Preliminary Assessment and Economic Evaluation of the Cauchari-Olaroz Lithium Project, Jujuy Province, Argentina.

Notes

Información adicional

Clasificación JEL: Q25, Q57

Información adicional

redalyc-journal-id: 5116