Lithium Mining in the Salar de Atacama—Accounting Practices for Water Footprinting

Abstract

As lithium demand increases, lithium brine from hyper-arid salt flats is becoming increasingly important, although there are concerns about its extraction’s impact on the local water balance. Water footprinting could address these impacts, yet studies lack consensus on whether to classify brine as water or a mineral. This study aims to review perspectives on lithium brine accounting within and beyond the water footprint context, focused on the Salar de Atacama, Chile, and to establish accounting principles for water footprinting, following the relevant ISO standard. Outside water footprinting, four perspectives on brine classification are identified: hydro-social/perceptual, molecular–thermodynamic, precautionary, and legal. Adopting some of these perspectives, e.g., the rationale of brine’s molecular–thermodynamic similarity to freshwater, some water footprint studies argue for accounting brine as equivalent to freshwater. However, they are a minority. According to ISO, brine should not be classified as freshwater, and the type of water and its functionality should be distinguished. We suggest some saline waters below a specific salinity threshold may function as freshwater and could be included in freshwater accounting. Additionally, lithium brine extraction can induce effects on surrounding water compartments. Since conventional water footprints overlook such local effects, we propose testing a set of site-specific accounting principles through case studies.

1. Introduction

1.1. Background

Lithium is of great importance for product performance in various applications, most notably in batteries, due to its high energy density and long lifespan [1]. It plays a central role in achieving decarbonization targets and driving the transition to clean energy [2], particularly through the electrification of the transport sector [3,4]. Due to potential supply risks and its high economic importance [5], lithium is a critical resource [6,7,8] with progressively growing global demand in recent years [9]. The growth in demand is primarily due to its increased use in batteries for portable electronic devices, electrical energy storages, and electric vehicles, which represent the largest end-use for high-grade lithium products since 2015 [10]. Favored by these conditions, global lithium production reached a peak of about 240,000 tons per year in 2024 [11], with future demand expected to continue to grow steadily until 2040 [12]. To date, the number of known lithium reserves totals 115 million tons worldwide [11], with the types of primary lithium reserves being generally divided into either ore-based resources such as spodumene minerals [13,14] or brine-based sources [15]. The main lithium reserves from ores are found in Australia (8.9 million tons), which is the largest lithium-supplying country globally, with 88,000 tons of lithium mined in 2024 [11]. However, the majority of the world’s known lithium reserves (approx. 60%) are found in continental brines [16,17]. Most of it is in the countries that belong to the region known as the Lithium Triangle [18], which includes Bolivia (23 million tons), Argentina (23 million tons), and Chile (11 million tons) [11]. Vast salt flats are located there, where lithium deposits are mined from large underground brine reservoirs [18,19]. Lithium brines can have relic or fossil characteristics and have formed over thousands of years, favored by regional climate conditions [20,21]. Considering the mining process, the lithium brine is pumped out and further concentrated in evaporation ponds under the influence of the sun’s intense radiation before it is sent for chemical processing [22]. Due to the low production costs (1,800 USD/ton lithium carbonate equivalent compared to production costs of 5,000 USD/ton for ore-based production in Australia [23]) and the highest known average lithium content (⊘ = 0.14%), the Chilean region is of particular importance for meeting current and future global lithium demand [24]. In this context, the Salar de Atacama (SdA) in northern Chile represents the largest producing brine deposit in the world. Its favorable conditions are due to the geological composition of its deposits and ideal climatic conditions characterized by high evaporation rates in a hyper-dry continental basin [24,25]. Taking into account current production, the global market share of lithium originating from the SdA is about one third [11] and is expected to remain in that scale in the long term due to the enormous reserves there and projected lithium demand increase [26]. However, this also raises the question of the potential environmental impacts of the current and future mining operations on-site.

The SdA forms part of the Atacama Desert, known as the driest and oldest desert in the world [27,28,29]. The main features of the natural water balance of the basin are determined by evaporation and precipitation, with no relevance of discharge to the sea [30,31]. The SdA is home to particularly unique and valuable ecosystems, including the Reserva Nacional de los Flamencos, an official Ramsar site, which hosts a variety of different lagoon ecosystems [31]. Vitally important for the reproduction of the high Andean flamingos, these ecosystems are likewise essential for further emblematic species, as well as for numerous extremophile organisms occurring in the salt flats and lakes [32]. In addition to the fauna, occasional vegetation assemblages are of additional significance to the unique biodiversity of the SdA [33,34,35].

There have been increasing concerns about the potential negative impacts of lithium brine mining in the SdA, including effects on the hydrological balance, the spread of droughts, salinity, biodiversity, and local communities [36,37,38]. This relates to the point that lithium brine mining could result in a lowering of local groundwater levels, less water being discharged into local lagoon systems, or a potential seepage of less saline groundwater into the brine aquifer [39,40]. These issues have been discussed in terms of their impact on local fauna, including bird species such as flamingos that rely on lagoon systems [37]. Additionally, the literature has hypothesized a potential relationship between lithium brine mining, reduced water availability for local agriculture, and lower agricultural production in the SdA [38,41].

In theory, these concerns could be assessed using methods such as the water footprint. There are two methodological approaches to the water footprint, the Water Footprint Network approach [42] and the life cycle assessment (LCA) approach based on an ISO standard [43]. Studies following the Water Footprint Network most often rely on volumetric assessments, taking into account blue water (surface and groundwater use), green water (rainwater use, provided it does not contribute to runoff), and grey water (an indicator of freshwater pollution, representing the volume of water required to dilute pollutants to an acceptable quality level) [42]. They are limited to “those human activities that impact upon the quantity or quality of freshwater within a catchment or river basin” [42]. On the other hand, the ISO water footprint [43], based on LCA [44], is defined as an impact-oriented measure, quantifying the water-related impacts throughout a product’s life cycle. It is best suited to practitioners seeking ready-to-use impact assessment for analyzing global product systems. It uses volumetric water inventories linked to the product system—such as the amount of water consumed—and combines them with regionalized impact assessment methods [43,45]. However, the method lacks standardized techniques for evaluating the complex local geohydrological dynamics in extreme environments such as the SdA. Furthermore, a more fundamental challenge exists at the core of any LCA-based water footprint assessment for lithium mining: there is currently no consensus on how to account for the extraction of brine on a volumetric level, i.e., as water or a mineral. For instance, an LCA study by Kelly et al. [46] accounted lithium brine and freshwater abstraction separately from each other. A study by Schomberg et al. [47], on the other hand, created a water footprint of a lithium-ion battery storage system and treated the total amount of lithium brine extracted in the SdA as water consumption. While lithium brine in Chile is legally considered a non-concessional mineral [48], the question of whether lithium should be accounted for as a water resource or a mineral resource is also influenced by the perceptions of different stakeholders and different perspectives on the issue [38,48,49,50]. Therefore, there is a need to compare existing case studies and concepts to develop consistent recommendations for brine accounting in the water footprint. This will allow for a more robust assessment of the trade-offs between the environmental benefits of lithium brine mining in terms of decarbonization potential and its water-related impacts.

1.2. Goal

This work aims to provide clarification on how brine should be treated and accounted for in water footprinting. The objective of the study can be classified into two sub-goals. First, current accounting practices for the consumption of brine, both within water footprinting and beyond, are summarized and compared, along with associated inconsistencies. This is done through reviewing the relevant literature and identifying if brine is accounted as a mineral or as water. The focus is on studies in the spatial context of the SdA, the most relevant location for lithium production from brine at present. Second, a conceptual framework on how to account for the consumption of brine in salt flats is developed. For this purpose, accounting principles for both the consumption of brine as such and the induced effects on the surrounding water bodies will be considered. Furthermore, saline waters of lower salinity than brine are discussed. The results of this work are primarily aimed at deriving recommendations on how to include brine in the water inventory of a water footprint, as a basis for determining potential impacts.

2. Material and Methods

2.1. Current Accounting Principles and Associated Inconsistencies

To address the first sub-goal, a total of 18 publications were reviewed, covering water footprint applications and studies that address the question of whether brine should be considered as water or mineral beyond the water footprint context. First, the studies outside the water footprint context and their position on how lithium brine should be considered were compared, as some of their rationales were relevant to the water footprint studies. This referred to seven peer-reviewed studies, grouped according to the general perspective on the question of whether brine is considered a mineral or a water resource [38,48,49,50,51,52,53]. Perspectives on brine accounting refer to the different viewpoints that shape how brine is understood and categorized in different contexts. For instance, a hydro-social perspective emphasizes the views of local and indigenous communities, highlighting their relationship with brine and water in salt flat ecosystems. The review of the relevant literature was conducted in a way that led to the definition of this and other overarching perspectives, providing different lenses through which brine can be interpreted. For each of the defined perspectives, the associated core reference(s) were provided. Additionally, the main conclusions on whether lithium brine should be classified as a mineral or a water resource were summarized.

The remaining eleven relevant, peer-reviewed publications refer to studies classified as water footprint applications, with a spatial focus on the SdA, but partially covering other salt flats [40,46,47,54,55,56,57,58,59,60,61]. These also include LCA studies that specify their water life cycle inventories and/or associated water-related impact assessments.

The water footprint and LCA-based related literature were grouped into the following categories:

• Studies that account for brine as equivalent to freshwater;
• Studies that equally consider both (i.e., including and excluding brine in freshwater consumption);
• Studies that do not account for brine as equivalent to freshwater.

Additional information was compiled in a table summarizing the study titles, the rationale for the different ways of accounting for lithium brine consumption, whether these studies accounted for the induced effects of brine consumption on the surrounding environment, and whether they made general statements about the ecological role of brine in salt flat ecosystems and possible ecosystem impacts of brine consumption. Then, a concluding summary was made of the rationale for accounting for brine in water footprint studies using different accounting principles.

Finally, all the identified literature was discussed based on the outcome of the second sub-objective, the derivation of volumetric accounting principles for the consumption of brine in salt flats.

2.2. Deriving Principles for a Conceptual Accounting Framework

The accounting principles for lithium brine in the water footprint were developed based on the ISO 14046 standard for LCA-based water accounting [43], which was referenced in ten [40,46,47,55,56,57,58,59,60,61] of the eleven water footprint studies previously reviewed. In the following, the methodological steps are described:

• ISO 14046 [43] was reviewed against the relevant definitions for brine, freshwater, and water of other salinities. This was used to determine whether or not the ISO allows brine to be accounted as equivalent to freshwater.
• Considering the different salinity and therefore water quality of brine and freshwater [62], it was then analyzed what the ISO standard says about how water quality should be considered and whether additional pros and cons for a specific accounting approach can be derived from this.
• Third, it was considered whether the standard specifies what salinity levels can and cannot be considered as freshwater and whether the standard’s definition of freshwater applies to regions with higher background concentrations of salinity, such as the SdA.
• Fourth, the standard was reviewed in relation to the inclusion of saline water in water scarcity-based assessments.
• Fifth, considering the possibility of induced effects of brine extraction on the surrounding environment [36,37,39,40], the extent to which the standard specifies how interactions with other water compartments should be covered was analyzed. Based on this and known concerns about lithium brine mining reported in the literature, the types of interactions that can be included at the water inventory level were identified.
• Finally, it was suggested which accounting principles should be applied in the context of lithium brine mining, taking into account a comprehensive assessment according to ISO 14046 [43].

Although we have tried to follow the wording of the standard as closely as possible, in some cases, some interpretation was necessary to derive accounting principles based on ISO 14046 [43].

3. Results

In the following sections, we present the results for the two sub-goals of this work. Section 3.1 and Section 3.2 summarize the findings from our review of relevant literature on brine accounting principles, both outside and within the water footprint context. Section 3.3 then presents the results of our proposed accounting principles, based on ISO 14046 [43], for LCA-based water footprinting.

3.1. Accounting Implications When Considering Studies Outside Applied Water Footprinting and LCA

Before moving to the comparison of water footprint and LCA studies (Section 3.2),

Table 1. Literature identified outside the water footprint and LCA context on the question of whether lithium brine should be considered as water or mineral.

Perspective or Viewing Angle	Studies	Conclusion on Whether Lithium Brine Should Be Considered a Mineral or Water
Perception-based or hydro-social perspective	-Based on Jerez et al. [38]: “Lithium extractivism and water injustices in the Salar de Atacama, Chile: The colonial shadow of green electromobility” -Also addressed by other studies, e.g., by Liu and Agusdinata [51], Lorca et al. [52], and Ramos Chocobar and Tironi [53]	Brine is considered to be water. Brine and freshwater are linked in the indigenous worldview and have a sacred character.
Molecular–thermodynamic perspective	-Based on Ejeian et al. [49]: “Is lithium brine water” -Supported by Schomberg and Bringezu [50]	Brine is considered to be water because of its molecular and thermodynamic properties, which are similar to those of pure water.
Precautionary perspective	-Based on Schomberg and Bringezu [50]: “How can the water use of lithium brine mining be adequately assessed?”	Unless there is strong evidence that the effects of lithium brine mining are negligible, the water scarcity footprint should take into account every cubic meter of brine consumed.
Perspective on legal terminology and its implications on water governance structures	Based on Flores Fernández and Alba [48]: “Water or mineral resource? Legal interpretations and hydrosocial configurations of lithium mining in Chile”	The current legal status quo in Chile treats lithium brine as a mineral resource and not as water. However, alternative interpretations are not only possible but also legally sound. These relate to a hybrid approach and the consideration of brine as a type of water.

shows the literature identified outside the water footprint and LCA context on the question of whether lithium brine should be considered as water or mineral. The studies were grouped according to their main perspective or viewpoint on brine accounting, and a total of four perspectives were defined, relating to a perception-based or hydro-social perspective [38,51,52,53], a molecular–thermodynamic perspective [49,50], a precautionary perspective [50], and a perspective on legal terminology and its implications for water governance structures [48]. Each of these perspectives is explained in more detail below, and the studies that address or adopt them are linked.

The study by Jerez et al. [38] is identified as the main work adopting the perceptual or hydro-social perspective. Jerez et al. [38] analyzed the linkage between electromobility, lithium extraction from the SdA, and its potential effects on the local indigenous communities. Regarding the question under consideration, they referred to the hydric nature of brine and criticized the viewpoint of accounting for it as a mineral resource. They underlined their argumentation with the statement that a consideration as mineral “does not acknowledge its hydrogeological complexity nor indigenous uses and world views of the Salar” [38]. The key part of their work focused on perceptional or hydro-social viewpoints. In this context, the authors argued that recognizing brine as a mineral resource neglects the indigenous communities’ cosmovision of brine as water with sacred character, directly connected to the land and people. In addition to the work of Jerez et al. [38], there are other studies that partially address the hydro-social point of view. One example refers to a work by Liu and Agustinata [51], who considered multiple dimensions of lithium mining stress in the SdA, including the perceived stress of local community members and their indigenous culture or identity. Lorca et al. [52], on the other hand, analyzed mining in the SdA in terms of potential consensus, tensions, and ambivalence of local communities. They highlighted that local communities are generally concerned about water scarcity issues linked to lithium mining, and that some share the view that all water is one, forming part of a living being with complex connections to soil, plants, and animals. However, the authors also stressed ambiguities and the difficulty of defining a single indigenous identity or consensus in a changing environment [52]. Additionally, Ramos Chocobar and Tironi [53] described the traditional Atacameño indigenous perspective in the SdA basin. They considered the water evaporated from lithium ponds and stressed that brine extraction in the salar affects the complex system connecting humans, ancestors, and land, changing the hydrological equilibrium of the SdA [53].

The second perspective on lithium brine accounting shown in Table 1 refers to the molecular–thermodynamic perspective with the associated core reference provided by Ejeian et al. [49]. Ejeian et al. [49] captured the core question of interest right in the title: “Is lithium brine water”? To answer the question, the authors applied an approach based on molecular dynamic simulations. SdA brine was simulated as well as pure water at five different pressures. Various molecular properties such as H-O-H bond angle and length, average entropy values, radial distribution and number of hydrogen bonds, and surface charge densities were investigated and compared between the different simulation products. Additionally, thermodynamics were considered while calculating Gibbs free formation energies regarding the brine. The results of their analysis showed that the water molecules of the brine behave similarly to pure water at 1.2 atm pressure. In addition, a Gibbs’ formation energy analysis revealed that more than 99% of the brine’s formation energy comes directly from water instead of dissolved minerals. Ejeian et al. [49] concluded, based on their findings, that “just as pressurized water is still called water, so is brine, since there is no thermodynamic phase demarcation between the two”. This interpretation is also found in the work of Schomberg and Bringezu [50], who stated that “molecular dynamics simulations strongly indicate that lithium brine is a type of water”, and who are linked below to a third additional accounting perspective.

Schomberg and Bringezu [50] wrote a position paper on how water use of lithium brine can be addressed. They refer to brine as water and base this, along with its material and molecular dynamic properties, on a rationale we call the precautionary principle. The main rationale behind it is that lithium brine extraction and evaporation can potentially affect the hydrological regime of the salars and surrounding ecosystems [37]. They therefore propose to take into account all human-induced losses from water bodies that may affect the natural availability of freshwater, including inland saline water. Every cubic meter lost in relation to local availability should be accounted for, regardless of whether it is brine or freshwater. They argue that this procedure should be followed unless it can be demonstrated that the environmental impact of brine extraction is negligible below certain extraction thresholds [50].

Finally, the fourth perspective on whether brine is water or a mineral relates to the context of legal terminology and its impact on water governance structures, with Flores Fernandez et al. [48] providing a core work on this. Flores Fernández and Alba [48] analyzed a series of legal interpretations of lithium brine mining in Chile and how they shape the water governance structures of local salt flats. The authors intended to emphasize that the current legal interpretation of lithium brine as a mineral is open to question and that other definitions (lithium brine as a hybrid of mineral and water; lithium as a type of water) may be legally sound. Then, differences in practical implications were derived in relation to each legal interpretation. It was concluded that the definition of brine is “a highly political process” affecting distributional (e.g., related to extraction rights), procedural (e.g., related to inclusion of indigenous communities in decision-making), and recognitional justices (legitimization of political, social, and cultural standings).

3.2. Current Aaccounting Practices for the Consumption of Brine Within Water Footprinting

Table 2. Comparison of accounting principles for the consumption of lithium brine in water footprinting and LCA. With the exception Díaz Paz et al. [54] using the Water Footprint Network guidelines [42], all studies follow the ISO-based water footprint [43] or LCA [44] approach.

Studies that Account for Brine as Equivalent to Freshwater	Rationale for the Accounting	Accounting of Induced Effects	Statements on Brine’s Ecological Role and Possible Extraction Impacts
Schomberg et al. [47] “Extended life cycle assessment reveals the spatially-explicit water scarcity footprint of a lithium-ion battery storage”	-None -Reasoning elaborated in Schomberg and Bringezu [50]: based on precautionary principle and molecular dynamics	No	-Potential effects on regional water scarcity
Mousavinezhad et al. [58] “Environmental impact assessment of direct lithium extraction from brine resources: Global warming potential, land use, water consumption, and charting sustainable scenarios”	-In conventional brine systems, a significant amount of water is lost to evaporation	No	-Declining aquifer levels, of particular concern in water-scarce regions
He et al. [60] “Environmental and life cycle assessment of lithium carbonate production from Chilean Atacama brines”	-Water contained in the brine is considered as a water resource affecting the local water balance -Need for cautious accounting in such arid environments	No	-Brines are a part of the overall water balance of the system
Studies that Consider Both, Including and Excluding Brine in Freshwater Consumption	Rationale for the Accounting	Accounting of Induced Effects	Statements on Brine’s Ecological Role and Possible Extraction Impacts
Mas-Fons et al. [57] “Carbon and water footprint of battery-grade lithium from brine and spodumene: A simulation-based LCA”	-Rationale for including brine: brine extraction causes freshwater seepage into the brine aquifer -Rationale for excluding brine: brine is unfit for human consumption or agricultural use	Yes	-Statements focus on the impact of extraction causing freshwater seepage into the brine aquifer
Studies that Do Not Account for Brine as Equivalent to Freshwater	Rationale for the Accounting	Accounting of Induced Effects	Statements on Brine’s Ecological Role and Possible Extraction Impacts
Kelly et al. [46]: “Energy, greenhouse gas, and water life cycle analysis of lithium carbonate and lithium hydroxide monohydrate from brine and ore resources and their use in lithium ion battery cathodes and lithium ion batteries”	-Due to brine’s high salinity and limited direct human use	No	-Brine, along with freshwater, is called a key factor within salt flat ecosystems
Schenker et al. [55] “Regionalized life cycle assessment of present and future lithium production for Li-ion batteries”	-Brine is not directly used by ecosystems or humans as a water source	No	-Brine pumping can affect the hydrogeological systems with wetland and lake ecosystems
Chordia et al. [40] “Life cycle environmental impacts of current and future battery-grade lithium supply from brine and spodumene”	-Brine is accounted separately due to its higher salinity	No	-Possibility of freshwater seepage into saline aquifers, which should be counted as consumption if quantifiable
Khakmardan et al. [56] “Comparative Life Cycle Assessment of Lithium Mining, Extraction, and Refining Technologies: a Global Perspective”	-None	No	-Brine system may play a role for the local groundwater tables
Lagos et al. [59] “Carbon footprint and water inventory of the production of lithium in the Atacama Salt Flat, Chile”	-Brine not included in ISO 14046 -Brine cannot be used for human consumption, agriculture, and even less in ecosystems.	No	-No core messages about the ecological role of brine or the potential environmental impacts of its extraction
Marinova et al. [61] “Water footprint of battery-grade lithium production in the Salar de Atacama, Chile”	-High mineral content of the brine makes it unfit for human consumption -According to ISO 14046, brine is not freshwater	No	-Environmental risks from hydrodynamic interaction with nearby systems
Díaz Paz et al. [54] “The water footprint of lithium extraction technologies: Insights from environmental impact reports in Argentina’s salt flats”	-Brine is accounted separately due to its higher salinity	No	-Brine is an integral part of salt flat ecosystems; its extraction may impact biodiversity, disrupt ecosystem services, and cause freshwater aquifer salinization

shows the comparison of the accounting principles for lithium brine consumption of the eleven studies classified as water footprint studies. It can be seen that a total of three studies accounted for brine as equivalent to freshwater [47,58,60]. One study considered both the inclusion and exclusion of brine in freshwater consumption equally, without prioritizing one approach over the other [57], and seven studies did not consider brine as equivalent to freshwater [40,46,54,55,56,59,61]. Table 2 also shows the rationale for the inclusion or exclusion of brine, where available in the publications. Additionally, it indicates whether the study considered the induced effects of brine extraction on the surrounding environment, with only one doing so [57]. Finally, Table 2 summarizes the key statements made in the included publications on the ecological role of brine and the potential impacts of its extraction. The following subsections provide a detailed overview of the studies, grouped according to how they accounted for brine, and conclude with a summary of the main rationales and assigned importance of brine in salt flat ecosystems.

3.2.1. Studies That Account for Brine as Equivalent to Freshwater

The first water footprint study in Table 2 that accounts for brine as equivalent to freshwater refers to a study by Schomberg et al. [47], who applied a spatially-explicit water scarcity footprint of a lithium-ion battery storage based on LCA. The origin of the lithium was not exclusively limited to the SdA but referred to the global lithium production mix. Evaporation losses were by far the largest contributor to water scarcity, accounting for almost 90% of the water scarcity-weighted footprint. In this context, lithium brine extraction in the SdA was a major contributor, influenced by the prevailing water scarcity there. The result is influenced by the fact that brine was treated the same as freshwater. Schomberg et al. [47] considered the mining of brine not as mineral extraction but as water abstraction, without distinguishing the different characteristics and functions of brine related to its high salinity. Concerning the impact assessment of water scarcity-related effects, they applied the methodology Available WAter REmaining (AWARE), the current consensus method for performing an LCA-based freshwater scarcity footprint [63]. The rationale for treating brine as freshwater was not explicitly provided but can be inferred from the opinion paper by Schomberg and Bringezu [50], presented in the previous subsection, which discusses precautionary assumptions and molecular dynamics as justifications for this approach.

Mousavinezhad et al. [58] conducted an environmental impact assessment using LCA to evaluate direct lithium extraction (DLE) from brine resources compared to alternative production pathways, assessing the impacts in three categories: global warming potential, land use, and water consumption. DLE technologies eliminate the need for open air evaporation ponds, reducing evaporation losses of water from the lithium brine [58]. This emerging technology, analyzed in the spatial context of a pilot project in Nevada (USA), was compared with lithium production from brine using the conventional method in evaporation ponds (SdA (Chile) and Cauchari (Argentina) salt flats) and lithium mining from hard rock deposits in Australia and China. As in Schomberg et al. [47], water-related impacts were analyzed using the AWARE methodology [63] for addressing freshwater scarcity impacts. The authors accounted for brine as equivalent to freshwater because in conventional brine systems, a significant amount of water is lost to evaporation. This assumption, together with additional indirect water consumption, resulted in the highest volumetric water footprints for conventional brine-based production [58]. On the other hand, the authors concluded that both ore-based production and DLE technology also consume significant amounts of water, albeit less. For the DLE technology, for instance, this was mainly in the form of indirect water consumption (e.g., for the production of reagents, diesel, and electricity), while assuming zero evaporation losses from the extracted brine in the foreground system. In terms of the final water scarcity footprint, conventional brine-based production in the SdA emerged as the most critical pathway due to the high water scarcity potential of the region.

Finally, He et al. [60] conducted an LCA of lithium production from Chilean brines of the SdA. In addition to characterizing the potential environmental impacts using the impact categories from the ReCiPe LCA methodology, a volumetric and a water scarcity footprint were determined. The water contained in the brine was considered as a water resource for the water balance calculations based on the rationale that it forms part of the total water balance of the system. In addition, the authors argue that Chile’s high water scarcity calls for cautious consumption of water resources in arid regions, especially in the Atacama Desert, where water scarcity is higher than the national average. Volumetric and water scarcity footprints were calculated for different modeling approaches of local evaporation. They concluded that while the evaporation process is not the most significant process in terms of volume of water consumed, when weighted by regional water scarcity, it becomes very significant in all scenarios. In analogy to Schomberg et al. [47] and Mousavinezhad et al. [58], the water scarcity footprint was determined using the AWARE model [63].

3.2.2. Studies That Consider Both, Including and Excluding Brine in Freshwater Consumption

Of the studies compared, one study used both accounting methods, including and excluding brine in freshwater consumption, without prioritizing one approach over the other. This refers to Mas Fons et al. [57], who conducted a comparative carbon and water footprint of battery-grade lithium from brine in the SdA vs. from ore in Greenbushes, Australia. As Table 2 shows, the rationale for including brine extraction for the water footprint is that it can cause freshwater seepage into the brine aquifer and therefore results in a loss of freshwater in the system. The amount of freshwater consumed in this scenario is equal to this seepage flow. As the authors do not specify how this flow was determined, and given the complex hydrology of such mixing effects [39,40], we assume that it is most likely that every cubic meter of brine consumed corresponded to an equivalent inflow of freshwater into the brine aquifer. Using this rather conservative assumption, this study is the only one to consider the induced effects of brine extraction through these mixing effects. The rationale in the scenario for excluding brine, on the other hand, was that brine is unfit for human consumption or agricultural use due to its high salinity. The results of the study showed that when analyzing different grades of lithium deposits, the water consumption volumes for all grades were approximately doubled when brine was included in the water consumption calculations. In addition, for high-grade lithium deposits, the way brine is accounted for determines whether brine-based (if brine is not accounted for as water) or ore-based production (if brine is accounted for as water) requires less water per unit produced. However, when considering the water scarcity-weighted footprint, ore-based deposits had a lower impact regardless of the brine accounting approach due to the high regional water scarcity in the SdA.

3.2.3. Studies That Do Not Account for Brine as Equivalent to Freshwater

Seven studies in Table 2 do not consider brine as equivalent to freshwater and are presented below, highlighting their main rationale where available. The first study by Kelly et al. [46] represents an LCA analysis of lithium carbonate and lithium hydroxide monohydrate from brine compared to the equivalent products based on ore resources. The assessment considered water and energy life cycle inventories as well as potential greenhouse gas emissions. Regarding the mining from brine resources, the study set the geographic scope to the SdA. Kelly et al. [46] considered aspects of water salinity and the fact that the brine in the salar has ten times more dissolved solids than seawater. As a consequence, this type of water is neither suitable for agriculture nor for domestic use, and was therefore excluded from freshwater accounting [46,64,65]. Their results indicated a significantly higher volumetric water footprint for ore-based production compared to lithium from brine resources.

Schenker et al. [55] conducted an LCA on lithium carbonate production while considering three existing and two planned lithium brine plants in the regional context of Chile, Argentina, and China. Regarding the selection of impact categories, effects of climate change, particulate matter on human health, and water scarcity were considered. Water scarcity-related aspects were evaluated based on the AWARE method [63] and showed its greatest effect on the water inventory in the arid region of the SdA. However, in line with Kelly et al. (2021), they did not include the consumption of brine at the water inventory level. Instead they stated that “brines are not directly used by ecosystems or humans as a water source and should, thus, not be considered when applying this LCA method” [55]. However, they suggested that the direct and indirect effects of brine pumping are a task for local hydrogeological work.

Chordia et al. [40] investigated the life cycle environmental impacts of current and future battery grade lithium supply routes from brine and spodumene. Regarding the brine-based production, salars in Chile (e.g., the SdA) and Argentina were the starting point. Various data sources were used, such as the LCA study by Kelly et al. [46] and the Ecoinvent database. Impact categories considered included climate change impacts, resource scarcity, freshwater ecotoxicity, and, for water-related impacts, the water consumption potential of ReCiPe and the water scarcity-weighted assessment based on AWARE. In the water-related assessment, they accounted for freshwater consumed. However, for the brine-based production, evaporation from lithium brine was recorded separately. In terms of freshwater consumption, Chordia et al. [40] suggested that lithium brine extraction could further reduce freshwater availability by inducing freshwater seepage into the brine aquifer and causing dilution effects. As brine extraction volumes far exceed direct freshwater consumption, this impact could be significant. However, due to the lack of correlation between brine extraction and freshwater intrusion, these effects were not quantified.

Khakmardan et al. [56] performed a comparative LCA to quantify the environmental effects of varying lithium production routes. In doing so, a brine-based production of lithium carbonate in the SdA was compared with different mineral-based production routes (spodumene (Australia and China), hectorite (Mexico), and zinnwaldite (Germany)). It was a cradle-to-gate analysis, and the analysis was limited to the global warming potential and the volumetric water consumption of the alternatives. About water consumption, the water that evaporates from the brine was not considered, and no rationale for this exclusion was specified. As a result, water consumption in the background system of the brine-based production became the dominant factor. In line with Kelly et al. [46], brine-based production resulted in a lower volumetric water footprint than ore-based production.

Lagos et al. compiled an LCA-based carbon footprint and water inventory of lithium production in the SdA. The rationale for not accounting for brine as equivalent to freshwater was that brine is not included in ISO 14046 [43] and that it cannot be used for human consumption or agriculture, much less in ecosystems. In terms of volume of water consumed, indirect water use (e.g., water for producing energy and auxiliary materials) was the largest contributor to the volumetric footprint.

Marinova et al. [61] conducted a cradle-to-gate water footprint of lithium extracted from the SdA. Based on an LCA approach, the associated water inventories were assessed for potential impacts of freshwater vulnerability and scarcity using the AWARE [63] and WAVE + [66] characterization models. The rationale for not considering brine as equivalent to freshwater is that its high mineral content makes it unfit for human consumption. It is also noted that including brine in freshwater consumption would not be consistent with ISO 14046 on water footprinting [43], which provides definitions of what water can be considered freshwater. Nevertheless, the authors acknowledge the potential importance of brine to the salt flat ecosystems and highlight the risks of hydrodynamic interaction with neighboring systems. The volumetric water footprint results are in line with Lagos et al. [59] and highlight the importance of indirect water consumption. However, the relevance of direct water consumption in the foreground system located in the SdA became more significant when considering the water scarcity-weighted results.

Finally, Díaz Paz et al. [54] estimated the water footprint of lithium extraction in Argentina’s salt flats, comparing the volumetric footprint of battery-grade lithium carbonate from two projects, Olaroz and Fénix, using different extraction technologies. Fénix operates as a hybrid system combining DLE with evaporation ponds, while Olaroz relies solely on conventional brine evaporation. The authors applied the Water Footprint Network concept [42] instead of the ISO 14046 standard [43], focusing on blue water resource assessment and separately analyzing blue water scarcity by comparing lithium projects’ water consumption to basin availability. While brine consumption is excluded from the water footprint due to its high salinity, they emphasized its ecological relevance and advocated for separate brine disclosure per unit of production. The analysis showed that the extraction process using DLE technology consumed more freshwater but less brine, whereas conventional evaporation had the opposite effect. This aligns with Vera et al. [67] and Halkes et al. [68], who highlight the need for sustainable freshwater alternatives in DLE operations. However, it contrasts with Mousavinezhad et al. [58], where DLE performed better due to brine being considered freshwater in the comparison. A potential water footprint component not included due to data limitations was the mixing of freshwater and brine due to mining. The authors noted that freshwater infiltration into the brine aquifer would increase the blue water footprint, while brine infiltration would contribute to a grey water footprint, indicating freshwater pollution.

3.2.4. Summary of Brine Accounting Rationales of Water Footprint Studies

After reviewing available water footprint and LCA studies, it has become evident that studies using similar accounting methods tend to share common rationales. Advocates for treating brine as equivalent to freshwater argue that, in arid environments, cautious accounting is essential due to its connection to the overall water system and potential negative impacts [47,50,60], such as groundwater table decline [58] and intensified regional water scarcity [47]. Additionally, the large volume of water lost to evaporation in conventional brine-based production was given as a reason to account for brine consumption as equivalent to freshwater [58] or the hydric nature of brine based on molecular dynamics [47,50]. On the other hand, studies that do not consider brine as equivalent to freshwater base their arguments on its high salinity, making it unfit for direct human or ecosystem use. It is important to note that none of the latter studies claim that brine consumption is without risk. In fact, the studies mostly agree that the brine system can be of significant importance to ecosystems and that its consumption can lead to potential local impacts. These induced effects, however, usually fall outside the scope of the studies compared.

3.3. Deriving Accounting Principles for the Extraction of Brine in Water Footprinting

Based on the review and interpretation of ISO 14046 [43], the resulting accounting principles for lithium mining are set out below.

3.3.1. Brine Cannot Be Considered Freshwater by Definition

The ISO water footprint standard does not explicitly refer to brine [43]. However, based on its definition of freshwater, which typically refers to water with a total dissolved solids (TDS) content ≤ 1000 mg/L and to water that is suitable for abstraction and conventional treatment to produce potable water, brine cannot be considered freshwater [43]. Brine is defined in the literature as water with a TDS above 100,000 mg/L [62], and SdA brine has TDS values of approximately 300,000 mg/L [69]. Water with such a high salt content, many times higher than seawater (30,000 mg/L) or brackish water (1000 to 30,000 mg/L) [43], is usually not used as a source to produce potable water.

3.3.2. Brine and Freshwater Should Be Reported Separately Based on Recommendations for Different Water Types, Qualities, and Functions

With regard to water quality, Section 5.2.4.1 (b–c) of the ISO standard [43] states that, among other data to be collected, the following water-related data should be considered for data collection:

“Types of water resources used”;

“Data describing water quality”.

About the categories of water resources that can be distinguished, Section 5.3.2 of ISO 14046 gives examples such as surface water, sea water, brackish water, groundwater (excluding fossil water), and fossil water [43]. Brine is not explicitly mentioned, but the fact that categories are given for different salinities (surface water vs. seawater vs. brackish water), but also between renewable and fossil groundwater, suggests that defining different categories for brine and freshwater would be consistent with the standard. In addition, Section 5.3.2 of the ISO [43] specifies that “information on each elementary flow should generally include, where relevant, water quality parameters and/or characteristics”, or “functional water quality descriptors”. Given the interconnected yet distinct roles of brine and surrounding water [31], there is a strong case for prioritizing water functionality analysis when assessing brine and freshwater in salt flats. The decision of which descriptors to use is a value choice. At a more general level, different types of quality descriptors are possible, covering functional characteristics from an ecosystem, human use, or resource perspective, but also intrinsic or perceived values. However, based on the priority of the scientific approach, the ISO states that “decisions within a water footprint assessment are preferably based on natural science” [43]. As brine and freshwater serve different functions in terms of human and ecosystem use, and also as a resource, our most appropriate interpretation of the standard is that brine and freshwater should be reported separately.

3.3.3. Accounting of Water Above the 1000 mg/L TDS Threshold

Given the TDS threshold for typically defining freshwater, most of the water abstracted in the SdA would not be considered freshwater due to the increased salinity of water there in general [41]. Therefore, it is also meaningful to consider what local authorities define as water for human use in saline environments. For example, in the context of the SdA in Chile, water with a TDS ≤ 5000 mg/L can be used for agricultural purposes [70] and that with a TDS ≤ 1500 mg/L for drinking water applications [71]. This suggests that some waters in the SdA below such local salinity thresholds but above the TDS threshold of 1000 mg/L may still function effectively as freshwater for human purposes. In such cases, it could be argued that they should be included in freshwater accounting or that the TDS threshold should be adjusted depending on the local conditions.

3.3.4. Addressing Saline Water Scarcity

When considering aspects of water availability, the standard mentions that pressures on types of water other than freshwater may be taken into account (ISO 14046: 5.4.5 [43]). With respect to saline water, the ISO provides the exceptional example of inland seawater, which can be subject to water scarcity [43]. However, according to the ISO [43] definition of freshwater, footprints that take into account highly saline water cannot be called freshwater scarcity footprints and thus cannot rely on methodologies designed for freshwater-related impacts (e.g., AWARE [63] and WAVE+ [66]).

3.3.5. Induced Effects of Lithium Brine Mining

In addition to accounting for the direct consumption of water of varying quality, the ISO water footprint standard also requires data to be collected to account for induced environmental impacts where relevant [43]. In Section 5.2.4.1, the standard says that the following data related to water shall be considered for data collection [43]:

“Changes in drainage, stream flow, groundwater flow or water evaporation that arise from land use change, land management activities and other forms of water interception, where relevant to the scope and boundary of the study being undertaken”.

In the context of brine extraction in the SdA, a number of induced effects of brine pumping have been discussed. These range from the lowering of the water table in the brine body and surrounding aquifers [37,72], to the reduction of groundwater recharge to local lagoons [36,37]. Other effects include the potential mixing of lower saline groundwater with the brine [39,40], and the effects on precipitation and recharge pulses in the region based on altered evaporation [73,74]. As some of these effects relate to the main concerns raised about lithium brine mining, investigating them may be highly relevant. Ultimately, however, the goal and scope phase of each study will have an influence on the water inventory items that need to be collected and on the impact assessment to be carried out.

3.3.6. Building the Basis for a Comprehensive Water Footprint Assessment Related to Lithium Brine Consumption

Depending on the scope of a water footprint study, according to ISO [43], the focus can be set to specific aspects while using qualifiers that describe a specific type of impact assessment (e.g., water scarcity footprint for evaluating the effects of freshwater consumption on regional water scarcity). However, a comprehensive assessment should “consider all environmentally relevant attributes or aspects of natural environment, human health, and resources related to water, including water availability and water degradation” [43]. What follows is a proposal for a comprehensive set of water accounting principles at the water inventory level to address the potential water-related impacts of lithium brine mining, which could form the basis for subsequent impact assessment models. Figure 1 shows a set of ten principles (a) to (j) that relate to lithium brine consumption.

On the basis of ISO 14046 [43], and in view of the inconsistencies in accounting practices to date, we suggest that the issue of water functionality (principle (a) in Figure 1) is of key importance. In this context, we argue that functional descriptors should be applied not only to the abstracted elementary flows themselves (e.g., the pumped brine) but also to all water compartments on which there are induced effects due to mining. For instance, if brine extraction results in a lowering of the water table, it is intended that this will be limited to the core mining area and will not affect the hydrogeological system in zone of sensitive ecosystems that rely on groundwater [73]. Given the impact endpoints commonly addressed in LCA-based approaches, the most methodologically appropriate functions in this context are those related to ecosystems (ecosystem quality endpoint), human well-being (human health endpoint), and resource value (resource endpoint) [44]. On this basis, although brine would be quantified separately from other types of water consumption, it would not be used as an inventory item to quantify the water deprivation potential of humans or the surface ecosystem. Instead, and due to its unsuitability for direct consumption, it is the induced effects of brine consumption on the overall water budget of the region that should be quantified. These include the artificial evaporation of lithium ponds (b) and the lowering of the average groundwater level (c) in certain spatial units of a study region [37]. The lowering of groundwater levels induced by lithium mining can lead to a reduction in phreatic evaporation in the study area (d), which may dampen further drawdown within certain groundwater depths (0.5 to 2 m) [73]. Marazuela et al. [73] termed the latter phenomenon the damping capacity of salt flats. In addition, the brine aquifer has the indirect function of inducing an upward flow of lower saline groundwater in the region, which then feeds local lagoon systems [31]. The aspect of altered groundwater discharge to the lagoons and changes in lagoon surface area due to mining [37] is represented by point (e) in Figure 1. Point (f) then relates to the potential for infiltration of lower salinity groundwater into the brine aquifer [39,40], which could indirectly deplete lower salinity water. The aspects (b) to (e) influence the total evaporative net water consumption on site (g). Net evaporative water consumption is defined here by summing the artificial evaporation from lithium ponds and potential reductions in evaporation due to declining groundwater tables and lagoon areas.

Finally, evaporated water is transported in the atmosphere to remote regions, but a certain fraction per basin usually rains down in the originating basin within a short period of time [66,75]. Berger et al. [66,75] termed the latter the basin internal evaporation recycling (BIER). The BIER has already been integrated into current water footprint methodologies [66,75,76] and represents an accounting principle that, to some extent, may influence basin recharge (h) in the SdA [74].

So far, the presented accounting principles address functional water aspects and the induced changes in water quantity (brine consumption, net evaporation, and precipitation feedback), surface area (lagoons), height (water tables), and quality (potential mixing effects). The next principle displayed in Figure 1 provides the link to the functional unit of the system under consideration (i). At this point, the brine consumed and any induced changes in other water compartments are linked to a quantified function of a product system [43]. For instance, this could be providing one ton of concentrated lithium brine, obtained after a certain residence time of the brine in the evaporation ponds, which serves as an intermediate product for further processing into battery-grade lithium compounds. Furthermore, we suggest separating the effects of brine extraction from other pressures within and outside the product system considered (e.g., by additionally considering the extraction of lower salinity water from wells outside the core mining area). In the case of possible co-products, allocation procedures shall take place [43]. The accounting principles presented attempt to reflect the induced water-related changes of brine consumption to date. However, due to the sensitivity of an ecosystem such as the SdA, additional future scenarios are recommended to minimize risks (j).

4. Discussion

4.1. Discussing the Brine Accounting Practices of the Literature Summarized

In the following, the literature reviewed in Section 3.1 and Section 3.2 will be discussed in more detail. In doing so, we first consider the more general literature outside the water footprint and LCA (Section 3.1). On the question of whether brine should be considered a mineral or water, we have presented seven different studies, with four different perspectives on the issue.

The perception-based or hydro-social perspective represented by the work of Jerez et al. [38] is influenced by the cosmovision of local indigenous communities. Local communities are protected by the national implementation of international conventions, such as Chile’s ratification of the Indigenous and Tribal Peoples Convention [77,78]. The latter specifies in Article 7 the right of indigenous communities “to decide their own priorities for the process of development as it affects their lives, beliefs, institutions and spiritual well-being and the lands they occupy, or otherwise use” [77]. However, we argue that this perspective is not applicable in the context of an LCA-based water footprint, which tends to focus rather on an environmental, human health or resource perspective [79]. In the context of LCA, perceptions and social impacts are addressed by the discipline of social life cycle assessment (S-LCA), for which two initial guidelines have been published [80,81], and the recently published ISO 14075 standard for product S-LCA [82]. An example of an S-LCA in the context of lithium mining in the SdA is the work of Roche et al. [83], which examines the associated social impacts with a focus on water-related potential impacts to the local communities.

The second perspective discussed here is the molecular–thermodynamic one introduced by Ejeian et al. [49], a perspective that is supported by Schomberg and Bringezu [50]. In principle, the design of the study is robust to show that, from a chemical and physical point of view, brine can be considered as a type of water. However, within the context of a water footprint, this is irrelevant, as these properties play no role for the impact pathways and have no link to the environmental impacts in the real world. This is easily demonstrated by the fact that seawater may be water as well under the molecular–thermodynamic perspective. By this logic, wherever a region borders one of the world’s oceans, water would be abundant. However, water scarcity footprinting usually focuses on available freshwater resources that can be used by local ecosystems and humans [63], excluding seawater. This example illustrates that the sole use of a molecular–thermodynamic approach is not suitable to answer the question of appropriate accounting for lithium brine in the context of (fresh)water scarcity.

Schomberg and Bringezu [50] proposed a precautionary perspective on lithium brine mining. While we acknowledge the use of such a perspective for testing the sensitivity of environmental accounting procedures in cases where system uncertainties are present, we strongly question whether this perspective should be used as the status quo for scientific assessments. We argue that a scientific environmental assessment and the models used in it should try to represent best the prevailing environmental conditions while reflecting the uncertainties involved. Decisions based on the scientific results may take a precautionary approach, but the scientific assessment as such should try to depict reality. In addition, there are two points of criticism regarding consistency when using such a perspective in the context of an LCA-based water footprint. First, in LCA-based approaches, it is not the exception but the rule that the description of the impact pathways for different impact categories is associated with uncertainties. If we intend to avoid a more differentiated assessment of anthropogenic pressures in water footprinting due to uncertainty, the question arises whether we want to extend this principle to all impact categories in LCA. Using theoretical worst-case approaches for environmental assessments does not support informed decision-making as the assumptions basically determine the results. Second, the precautionary perspective of treating different water types equally is not consistent with requirements of the ISO water footprint standard [43]. This states in Section 5.4.4.1 that the types of water resources (e.g., freshwater, brackish water, surface water, seawater, groundwater, fossil water) should be taken into account, when appropriate, for the characterization of potential impacts [43]. In the context of the SdA, we believe it is undeniable that, for instance, the depletion of freshly recharged groundwater or water from protected lagoons must follow different impact pathways and associated characterization approaches than the depletion of fossil underground brine. Hence, separate accounting is scientifically more robust.

The last perspective presented in Section 3.1 refers to Flores Fernández et al.’s [48] perspective on legal terminology and its implications for water governance structures. They reflect both explanatory patterns for considering one or the other terminology and detailed elaborations on the shaping of hydro-social relations and governance structures. While we believe this interaction is worth investigating from a political, social, and regulatory perspective, their work focusses more on stakeholder interests than science. In addition, their approach cannot be directly applied to LCA-based water footprinting, as this tends to be methodologically oriented towards the natural sciences, while highly politico-social issues are covered by other disciplines.

Finally, we move to the discussion of how brine is currently accounted for in water footprinting (Section 3.2). Seven out of the eleven reviewed LCA studies accounted brine separately from (fresh)water and did not include brine in their volumetric or water scarcity-weighted assessment [40,46,54,55,56,59,61]. We regard this in conformity with the current definitions of the ISO standard for freshwater [43] and, in terms of a subsequent water scarcity assessment, as most consistent with the AWARE method and its intended usage [63]. For a state-of-the art water footprint according to the ISO [43], the treatment of brine as a resource and not as water is appropriate. Impact-based water footprinting and LCA determine potential environmental impacts most often from a global perspective. They have generally limitations with regard to very specific and local impacts. None of the authors questioned the potential impacts of brine abstraction on the hydrogeological system and adjacent ecosystems, but these issues are subject to local hydrological research. If such local effects are relevant and if sufficient local hydrological information is not available, it is suggested that sensitivity analyses should be performed with these data. This is applied on a high level, for instance, in Chordia et al. [40], by contrasting the amount of brine with the amount of freshwater, hypothesizing on the potential local effects of the former. The studies by Schomberg et al. [47], Mousavinezhad et al. [58], and He et al. [60] were the only studies that prioritized accounting for brine and freshwater in the same way, without distinction. Their justification relies mainly on the molecular–thermodynamic [47] and precautionary principles [47,60] or the fact that a huge amount of the water content in the brine is lost due to evaporation [58]. As elaborated above, these criteria are inappropriate and not suitable for purposes in the water footprint context. In addition to this, the approach in the LCA study by Schomberg et al. [47] is inconsistent, as it combined the amount of brine abstraction with the AWARE method, a method specifically designated for freshwater scarcity footprinting and not for highly saline brine. The characterization factors of AWARE are not consistent with an elementary flow originating from brine.

4.2. Discussion of the Accounting Principles Suggested for Lithium Brine Consumption

Accounting principles on how to possibly consider the consumption of lithium brine have been presented in Section 3.3. This has led to recommendations on what accounting aspects need to be considered for a comprehensive water footprint assessment of lithium brine extraction in Section 3.3.6 (see Figure 1). It is important to note that these aspects require very localized analyses. At this level, LCAs are generally not performed by default. In the LCA context, a method is always more practical if it can be applied at global scale. However, it is questionable whether we will have operational methods at that level of detail that capture lithium brine mining across all relevant salt flats globally.

If an operational methodology is not available for the region of interest, we suggest the following as a first step: According to ISO 14046 [43], it is clear that brine is not freshwater. In that sense, it is not supposed to be accounted for. However, in the absence of alternative water sources, other waters above the official freshwater ISO threshold (1000 mg/L [43]) may effectively act as freshwater for human purposes. Thus, we recommend including such saline waters in the standard volumetric counting, e.g., based on salinity thresholds provided by local norms. Within the Chilean context, this could refer to waters with a TDS ≤ 5000 mg/L (threshold for water in agricultural use [70]) or a TDS ≤ 1500 mg/L (threshold for water used for drinking purposes [71]). The volumes of water consumed acting as freshwater can be combined with characterization factors describing potential freshwater scarcity-related effects. A suitable model for this would be the current consensus model AWARE [63]. While we support the exclusion of brine from the default volumetric accounting, its inclusion could be separately addressed in a sensitivity analysis if there is a large uncertainty about potential induced effects on surrounding environmental compartments.

As a second step, water accounting at a local level could be added to have a more comprehensive overview of the situation on-site, as the current ready-to-use water footprint methods do not reflect the specific conditions in salt flats. The proposed accounting principles in Section 3.3.6 could serve as guidance on what aspects should be accounted for, such as the lowering of groundwater table depths in zones that support local ecosystems. The accounting principles take into account the interactions between different water compartments, an approach supported not only by ISO 14046 [43], but also by other water footprint guidance documents (see, e.g., Núnez et al. [84]). In Nunez et al., the term “multi-media fate approach” was used to describe the quantification of the impact of extracting water from one compartment on those surrounding it. This approach was considered to be the next generation of water footprint indicators, succeeding those that focused on consumption-to-availability ratios. As the impacts of brine extraction primarily concern the induced effects on surrounding water compartments, increased research into localized water footprint analyses of lithium brine extraction could make these multi-media fate approaches more prominent in water footprinting.

Furthermore, the accounting principles cover aspects of water functionality. In the context of LCA and with a focus on ecosystem quality, biodiversity functions that affect local flora and fauna seem to be the most straightforward. In contrast, functions that consider extreme life forms within the underground brine reservoir, such as microbial communities [85], are typically excluded in applied LCA. Therefore, local accounting in the context of ecosystem quality should primarily address hydrological effects that could potentially affect higher plant and animal species (e.g., the reduction in surface area of lagoons that provide habitat for local wildlife). However, we emphasize that such an analysis should still be based on realistic, rather than theoretical, worst-case assumptions, while communicating the uncertainties involved.

In the absence of explicit accounting principles for highly saline water in official guidance documents, the suggested accounting principles for brine and other saline water could provide guidance and encourage the development of a consensus. We recommend stating the issue of accounting for brine-related effects in water footprinting more explicitly in the future revisions of the ISO water footprint standard [43] and other relevant documents, such as the guidelines from the Water Footprint Network [42]. The need for this was shown in Section 3.2 by the different accounting practices that currently prevent fair comparability.

5. Conclusions and Outlook

The first sub-goal of this work was to analyze current accounting practices for the consumption of brine, both within water footprinting and beyond, along with associated inconsistencies. The results of the literature review showed many perspectives on whether brine should be classified as water or a mineral and different approaches on how to account for brine in the water footprint. To remain consistent with the ISO standard for water footprinting [43] and the accounting principles derived from it, we conclude that brine should not be accounted for as freshwater, a practice that was followed by the majority of the water footprint studies analyzed. However, induced effects on surrounding water compartments through brine consumption (e.g., on nearby groundwater, local lagoons, phreatic evaporation, and precipitation feedback) are currently not sufficiently covered and could be considered in a localized water footprint analysis. Furthermore, in a region where there is no water that meets the 1000 mg/L standard defined for freshwater [43], some saline waters can still function as freshwater for humans and ecosystems. This functionality should be considered when accounting for water consumption in regions that do not have freshwater as classified by the ISO [43].

The accounting principles developed have led to a proposal for the type of water inventory information to be considered for a comprehensive water footprint assessment of lithium brine extraction. These represent localized analyses that need to be tested in future case studies and compared and evaluated according to their relevance. While these principles were developed with a focus on current mining practices in the SdA, they may also be relevant to other salt flats. Thus, in addition to assessing the relative importance of the accounting principles within the SdA, it is also important to assess whether the components of the framework are one-to-one transferable to other salt flat systems. This may include brine extraction in regions where lithium exploitation is not the focus and other elements such as sodium, potassium, or magnesium salts are more commonly targeted (e.g., Bonneville Salt Flats in Utah, U.S.) [86]. Moreover, this requires careful analysis, as salt flats can differ fundamentally in their formation history, salt concentrations, surface water connectivity, brine–freshwater interactions, and ecosystem functions [31,86,87]. Some salt flats, for instance, can be characterized by higher precipitation than in the SdA, leading to seasonal flooding and therefore different hydrological dynamics (e.g., Makgadikgadi pans in Botswana) [86]. For salt flats with different hydrodynamics than the SdA, different induced effects on the surrounding environmental compartments may be prioritized for investigation. It should also be noted that for different product systems, both within and outside the lithium context, other post-processing (after brine extraction) may take place that differs from the classical evaporation pond process [67]. In such cases, not all of the proposed accounting principles are relevant (e.g., principle (b) from Figure 1: artificial evaporation from ponds).

The accounting principles developed in this work can serve as a basis for promoting dialogue and developing harmonization guidelines for more consistent reporting of water-related impacts in lithium mining. Recently, an industry carbon accounting guideline specific to the lithium sector was introduced to improve the comparability of carbon accounting between different lithium production sites [88]. Similarly, in the long term, harmonized water accounting should be pursued as a strategic goal. This would enable the lithium industry to track the water footprint of its products in a standardized and transparent manner. Moreover, this work can be seen as an encouragement for companies to intensify their efforts in conducting localized case studies on the local environmental effects of lithium mining. The availability of data and models will be a critical factor and should be carefully assessed. The results of such case studies should provide a more comprehensive view of local conditions and help to identify potential tipping points where the hydrogeological system may be subject to long-term perturbations, taking into account the uncertainties involved. This should include sensitivity analyses, especially in relation to brine–freshwater interaction thresholds. Finally, if more studies focus on defining thresholds for brine extraction below which negative effects on local ecosystems can be ruled out in both the short and long term, this could be directly useful for policy. For instance, it could help local authorities make more informed decisions when issuing permits for brine extraction.