Exploring the Material Feasibility of a LiFePO₄-Based Energy Storage System

Abstract

This paper analyzes the availability of lithium resources required to support a global decarbonized energy system featuring electrical energy storage based on lithium iron phosphate (LFP) batteries. A net-zero carbon grid consisting of existing nuclear and hydro capacity, with the balance being a 50/50 mix of wind and solar power generation, is assumed to satisfy projected world electrical demand in 2050, incorporating the electrification of transportation. The battery electrical storage capacity needed to support this grid is estimated and translated into the required number of nominal 10 MWh LFP storage plants similar to the ones currently in operation. The total lithium required for the global storage system is determined from the number of nominal plants and the inventory of lithium in each plant. The energy required to refine this amount of lithium is accounted for in the estimation of the total lithium requirement. Comparison of the estimated lithium requirements with known global lithium resources indicates that a global storage system consisting only of LFP plants would require only around 12.3% of currently known lithium reserves in a high-economic-growth scenario. The overall cost for a global LFP-based grid-scale energy storage system is estimated to be approximately USD 17 trillion.

1. Introduction

The industrial revolution spurred an enormous increase in the use of energy in the latter half of the eighteenth century, and this increase has continued unabated ever since. Consequently, humanity has emitted copious amounts of carbon dioxide, an inevitable product of the combustion of fossil fuels—coal, oil, and natural gas—that have provided this energy. Although such combustion is harmful to human health at the localized, acute level, the long-term global consequences of carbon emissions are now being recognized. The primary concern is that of climate change driven by rising temperatures across the globe that are attributable to the anthropogenic carbon emissions. With average temperatures from 2011 to 2020 that were 1.1 °C above 1850–1900 levels [1], the Intergovernmental Panel on Climate Change (IPCC) estimates that the planet is on track to warm by at least 1.5 °C above 1850–1900 levels (and potentially more than that depending on the level of carbon emissions) in the coming decades [2]. This warming and consequent climate change has multiple effects, including heat waves, extreme weather, altered precipitation patterns, and sea level rise, that tend to degrade the planet’s ability to sustain healthy living conditions for both humanity and the natural world. Among the harmful manifestations of anthropogenic climate change are increased flooding in cities and population centers in coastal regions, rising food prices due to droughts, localized destruction of infrastructure, increased human and other species mortality, and so on [3].

A global consensus has been developing over the past few decades that a transition away from the current carbon-based energy system is essential to avoid altering the planet’s climate to such a degree that modern society will suffer from these expensive and deleterious effects. Energy decarbonization—the substitution of fossil fuels with clean (carbon emission-free) primary energy sources—will avert future carbon dioxide emissions and arrest or even possibly reverse global warming and climate change. Chief among fossil fuel-reliant infrastructures is the electric grid, with significant amounts of power generation and concomitant carbon emission coming from coal or natural gas plants. Renewable energy resources—mainly solar and wind energy—have the potential to successfully displace coal and natural gas as the primary energy sources for electricity generation. However, power generation from these renewable sources is inherently inconsistent due to variable conditions over time and often unpredictable weather patterns, and cannot be relied upon to provide uninterrupted and dispatchable electrical energy to large-scale grids. Grid-level energy storage systems, configured to charge during times of excess generation and deliver the stored energy during times of excess demand, are indispensable to a decarbonized energy future based on variable renewable energy [4]. Overall, achieving a 100% carbon-free grid with 100% reliability will require a technological solution to overcome the energy storage problem.

A possible solution for the energy storage challenge is through the large-scale deployment of lithium-ion batteries. The advantages of Li-ion batteries include their dispatchability, high energy density, low cost, relatively high round-trip efficiency, and modularity. However, Li-ion batteries are not without problems, specifically those revolving around scalability. The amount of energy storage required presents a significant challenge in the transition to a renewable-energy-dominated grid. Resource availability is a key concern in this area, with world lithium reserves being a potentially limiting factor for large-scale implementation [2].

Most Li-ion batteries in use currently are in electric vehicles (EVs) and other portable/mobile applications [5]. These batteries typically use nickel manganese cobalt (NMC) cathodes or other chemistries that include some combination of these metals [6]. NMC Li-ion technology is attractive in that its gravimetric energy density is higher than other types of Li-ion batteries, and the electric potential developed is acceptable for most applications [7]. These advantages are most beneficial for mobile applications/storage (e.g., vehicles and personal devices such as laptops) where weight and sustained performance are crucial for the success of the system. The suitability of NMC Li-ion batteries for mobile applications explains their widespread use in EVs. In stationary applications for electric grid backup, though, the energy density advantages are not as important. Indeed, NMC batteries’ reliance on scarce and expensive metals is a liability, especially when implementing the technology on as wide a scale as that required to decarbonize the electric grid. A comparison of various types of Li-ion chemistries is shown below in

Table 1. Li-ion chemistry advantages and disadvantages.

Criterion	Lithium Iron Phosphate (LFP)	Lithium Manganese Oxide	Nickel Manganese Cobalt (NMC)	Lithium Cobalt Oxide	Nickel Cobalt Aluminum	Lithium Titanate (LTO)
Cycle Life	High	High	Moderate	Low	High	High
Safety/Stability	High	High	Moderate	Low	Marginal	High
Specific Energy	Moderate	High	High	High	High	Low
Cost	Low	Moderate	High	High	High	High
Resource Concerns	No	Yes	Yes	Yes	Yes	Yes

[8].

The analysis presented in this paper is based on employing lithium iron phosphate (LFP) batteries for stationary electrical energy storage, avoiding reliance on cobalt and the potential ethical concerns associated with its extraction and use. The primary advantage of LFP batteries is that they do not require the use of expensive metals such as cobalt, manganese, or nickel [9]. These metals are scarce, and therefore expensive, with the possibility that the Earth cannot physically support the scale of the system required for global energy storage at the low cost required for wide-scale implementation. The specific concerns related to cobalt and manganese are further explained below.

Cobalt is produced primarily as a byproduct of nickel and copper extraction [10]. This introduces risk into the cobalt supply chain because the markets for nickel and copper have not grown as quickly as the cobalt market. This means that while demand for cobalt has risen, production has not, increasing the cost of cobalt and making it a limiting resource. Another source of risk to the supply of cobalt is the human structures and institutions surrounding cobalt extraction in the regions of sub-Saharan Africa, particularly the Democratic Republic of Congo (DRC), which as of 2017 controlled ~60% of the market. Extraction in the DRC is associated with environmental degradation, child labor, and political unrest [10], all of which induce risk into the cobalt supply chain [11]. Large-scale use of cobalt will likely exacerbate ethical concerns stemming from the systematic exploitation of ordinary people and expropriation of their collective mineral resources while degrading the environment with mine tailings and harming their health [12]. These concerns could be avoided through alternate sources of cobalt (for instance, deep-sea nodules—a source that raises other environmental concerns [7]), or utilizing other technologies that avoid cobalt usage altogether, as proposed through the use of LFP batteries in the current analysis.

Manganese, another metal critical in NMC-type Li-ion batteries, faces its own challenges. Principal among these challenges is the distribution of the resource. High-grade manganese ore is found only in a handful of countries, principally South Africa, Gabon, Australia, and China [13]. For other countries, including those with large future demand for manganese, this introduces a requirement to import large quantities of manganese ore to meet domestic demand. This in turn places the fate of a country’s clean energy transition into the hands of other countries which can restrict exports at any time. The challenge facing the manganese market is the combination of market risk, price volatility, and resource risk. The manganese ore market is one in which protectionism, lack of safety regulations, strong competition, and environmental degradation abound, which induces doubts about the sustainability of such a system. Additionally, large price swings occur due to the dominance of a few commercial entities, who are able to raise prices based on the annual contract system of supplying manganese. Finally, resource risk stems from the fact that world production of manganese has fallen short of world demand at times over the last decade [13]. While the supply chain risks of manganese are not as substantial as those for cobalt, they are still significant and should give policymakers pause before staking the future of energy storage on technology that depends on its use.

It is clear that both cobalt and manganese resources face risks in the future. Chief among these is the unequal global distribution of their ores, with significant reserves of both metals concentrated in countries with low political stability that can affect the supply of critically needed materials, slowing or even halting the transition to a decarbonized energy future worldwide. Another plausible scenario is that the production of these strategic metals could simply fail to keep up with demand, resulting in price spiraling and making the transition prohibitively costly.

LFP cycle life is higher than that of NMC-type Li-ion batteries [14], a major advantage in a utility-scale system. LFP also has high thermal stability compared to LiCoO₂, LiNiO₂, and LiMnO₂ [15], meaning that the battery thermal management systems (BMSs) in a grid storage plant can maintain the batteries at a higher temperature and reduce their energy consumption and size. The only real advantage NMC batteries have over LFP technologies is their gravimetric energy density, which is why NMC technology dominates in EV applications [14]. In a stationary energy storage system, this advantage is less relevant.

A potential alternative to Li-ion battery storage is Na-ion battery storage, based on the similarity of Li and Na battery chemistry. However, despite their similarity, significant differences exist between Li-ion and Na-ion batteries with respect to their structure and operational characteristics. Na-ion batteries require hard carbons due to their incompatibility with graphite due to the larger atomic size of Na that results in thermodynamic instability of graphite intercalated with Na. Further, Na is more active than Li, and the higher solubility and hydrophilicity of Na salts makes cathodes in Na-ion batteries highly susceptible to air and moisture attacks. The higher activity of Na also translates into lower stability of the solid–electrolyte interface layer and a tendency for dendrite formation. While the abundance of Na is three orders-of-magnitude higher than that of Li, Na-ion batteries suffer from almost a fourfold decrease in electrical energy storage capacity on a per-mass basis from Li-ion designs [16]. It should be noted that significant research and development activities are taking place for Na-ion batteries—for example, with respect to electrode materials as described by Xu et al. [17]—that may potentially advance Na-ion battery technology to make it competitive with other technologies. However, at the present time, the supply chains and manufacturing processes for Li-ion batteries are already extant, supporting vehicular and consumer electronic applications. This market presence is itself an advantage: rather than having to retool factories and re-engineer products, existing supply chains can simply be upscaled. The superior characteristics of Li-ion batteries (high specific energy and specific power, excellent round-trip efficiency, service life, low self-discharge, etc.) coupled with the advanced technology readiness level led to preferential consideration of Li-ion batteries over Na-ion batteries in the present analysis.

Overall, LFP batteries offer an attractive alternative for removing problematic metals—primarily Co and Mn—completely from the battery system, replacing them with iron and phosphorus, elements that have a much wider global distribution and larger amounts of reserves. These advantages outweigh any performance advantages of NMC-type Li-ion batteries. Therefore, the analysis presented in this paper assumes LFP to be the technology of choice for a global storage system. Estimated lithium resource requirement for a nominal carbon-free grid supported by LFP battery storage by the year 2050 is presented for both the reference case and a high-growth scenario. The resource requirement incorporates that for complete electrification of the transportation sector. An estimate of the total cost of the storage system is also presented.

2. Materials and Methods

To answer the question of whether there exists enough Li on the planet to support an LFP-backed carbon-free grid, two quantities must be estimated. First is the amount of energy storage required for sustained grid operations. This estimate is arrived at in this work from the grid makeup and the total grid size of the forecasted system in 2050. The second quantity that must be estimated is the amount of lithium needed per unit of energy stored. This quantity is estimated from the information available for a currently operating LFP energy storage plant. With these two quantities, it is possible to determine whether there exist sufficient mineral resources to pursue LFP as an energy storage option.

2.1. Configuration of a Carbon-Free Grid in 2050

2.1.1. Grid Composition

The 2019 world electric grid was composed as follows: 36.7% coal, 2.8% oil, 23.5% natural gas, 10.3% nuclear, 16.0% hydro, 0.5% geothermal and tidal, 5.3% wind, 2.6% solar, and 2.4% biofuels or waste, yielding a total electrical generation of 27,044 TWh [18]. To avoid the worst effects of warming (e.g., warming above 2 °C above pre-industrial levels) the UN calculated that the world must reach net zero emissions by 2050 [19]. To achieve this, all carbon sources of electricity need to be replaced by an energy source that emits no carbon.

This analysis assumes that the percentage of electricity generated by nuclear and hydro power stations will remain unchanged by 2050. Nuclear technology offers some obvious advantages such as extremely low resource requirements per unit of energy delivered, zero carbon emissions, and dispatchability. However, proven technologies, such as light water reactor technologies, have had difficulty in delivering new generation capacities on-time and within-budget, with the recent Plant Vogtle units in Georgia, USA, running seven years behind schedule and USD 17 billion over budget [20]. In addition, newer nuclear technologies, such as small modular reactors, have so far been difficult to implement and regulate, are likely to be expensive per unit of energy, and are overall unlikely to contribute significantly to the global electricity supply over the time horizon considered for this study. Although three designs are currently operational (a floating water-cooled reactor in Russia and two high-temperature gas reactors in China and Japan), the vast majority of SMR designs are in various stages of design and construction [21]. Over a 25-year horizon, it appears unlikely that the dozens of companies vying for regulatory approval and capital investment will be able to roll out enough SMRs to appreciably increase nuclear power’s share in the energy mix, particularly as energy needs continue to increase year on year. In addition, older plants will be shut down as time progresses, further decreasing nuclear power’s ability to grow as a percentage of the energy mix.

Hydroelectricity, too, seems unlikely to increase appreciably in the crucial next decades. This is because most of the suitable sites with the required flow/water and hydraulic head availability have already been developed. The potential sites where large-scale hydroelectricity projects can be located are expensive and faced with environmental concerns. Therefore, similar to nuclear power, this analysis assumes that hydroelectricity generation will also not increase appreciably by 2050.

The contributions of geothermal and tidal energy sources are not anticipated to grow significantly by 2050, and hence the contributions of nuclear, hydroelectric, and geothermal and tidal energy sources to the electric grid in 2050 are assumed to be same as that in 2019 at 2786 TWh, 4327 TWh, and 135 TWh, respectively [22].

The balance of the forecasted global electricity demand in 2050 is assumed to be satisfied through electrical generation made up of a 50/50 mix of solar and wind energy sources, eliminating all fossil fuel sources.

2.1.2. Forecasted Global Electricity Demand in 2050

The International Energy Outlook (IEO) of the US Department of Energy’s Energy Information Administration’s (EIA) (www.eia.gov) is a yearly publication that discusses possible future scenarios regarding global primary energy demand. The report details a myriad of factors that will impact future global energy demand. These factors include population growth, economic growth as measured in GDP per capita, and energy efficiency both at the industrial scale and residentially. The base case chosen for analysis is labeled the reference case in the data and assumes all current policies remain in effect [19]. Essentially, this case reflects the way in which the global energy system is trending if policy and technology costs remain the same.

Another adjustment to the EIA’s electricity demand forecast includes primary energy requirements for the transportation sector. Electrification of the transportation section is necessary to achieve a net-zero carbon future. While the path towards electrification of land transportation is understood well, the electrification of marine and air transportation sectors is more challenging. Nevertheless, carbon emissions associated with these sectors also need to be eliminated for a net-zero future. It is assumed that feasible technological solutions to electrification of these sectors will also be available for implementation by 2050, and the analysis presented below incorporates the forecasted demand of the entire transportation sector into the additional electricity requirement. To approximate these transportation energy requirements, the analysis uses the forecast demand for transportation primary energy and converts that value into an electrical value of additional TWh required to estimate the total electricity requirement in 2050. The EIA reference case posits that, in 2050, 42,298 TWh will be generated worldwide. Based on IEO 23, the projected energy consumption in 2050 including transportation energy consumption is shown in

Table 2. Projected energy consumption in 2050.

Scenario	Transportation Sector						Non-Transportation Sector TWh	Total TWh
	Jet Fuel		Gasoline		Distillate
	Quad	TWh	Quad	TWh	Quad	TWh
Reference Case	22.3	6535	47.6	13,950	49.4	14,478	42,298	77,261
High Economic Growth	26.2	7678	52.5	15,386	57.6	16,881	49,715	89,660

.

This yields a total electricity demand of 77,261 TWh for the reference case. The EIA also provides energy consumption predictions for 2050 in a high-economic-growth scenario, which yields a total reference electricity demand of 89,660 TWh. The overarching variable upon which the amount of required storage capacity depends is the percentage of generation that is not ‘firm’, e.g., the percentage of generation that is weather-dependent. Subtracting today’s nuclear and hydro generation from the projected electricity demand in 2050 gives the total generation from renewable sources. Of this value, this analysis assumes that half will come from solar photovoltaic (PV) and the other half from wind. To determine actual grid composition in MW of installed capacity, average solar and wind capacity factors were used—22.13% and 33.59%, respectively [23], calculated from US data from 2023. This solar capacity aligns with a 25% global capacity factor assumed by Lowe and Drummond in a global renewable analysis [24]. The postulated wind capacity factor is somewhat higher than the average in other countries such as China (0.22) and Germany (0.19) [25], but as offshore wind becomes more widespread, this number is expected to increase [26]. Therefore, 0.33 is a good estimate for the purposes of this analysis.

This results in a solar PV capacity of 18 TW and a wind capacity of 11.90 TW for the reference case. In the high-economic-growth case, the solar PV capacity and wind capacity are 21.25 TW and 14 TW, respectively.

As stated in Section 2.1.1, nuclear generation capacity and hydroelectric capacity will remain at 2023 levels (413 GW [27] and 1400 GW [28], respectively), meaning that the nominal grid makeup analyzed for the reference and high-economic-growth cases is as shown in

Table 3. Primary energy source contributions to global electric grid in 2050.

Scenario	Primary Energy Source	Grid Percentage	Nominal Power Capacity GW
Reference Case	Nuclear	1.3	413
	Hydro	4.4	1400
	Solar	56.8	18,045
	Wind	37.5	11,913
	Total	100	31,769
High Economic Growth	Nuclear	1.1	413
	Hydro	3.7	1400
	Solar	57.3	21,514
	Wind	37.8	14,192
	Total	100	37,545

:

It is important to note that these generation estimates are conservative in that they do not take into account any potential efficiency improvements in generation. The evolution in technologies will likely translate into a smaller requirement, with the result that estimation of lithium requirement based on these grid capacities is also conservative and adequacy of resources in the conservative limiting case will mean adequacy for all other cases.

2.2. Nominal Plant

The second step in determining the feasibility of a renewables-dominated grid backed by LFP technology is to choose a nominal LFP energy storage plant for the scale-up analysis. This plant should be one that is currently in operation, to avoid any overly optimistic projections for a plant that exists only on paper. The nominal plant selected for this analysis is a 20 MW/10 MWh capacity LFP-type energy storage power station serving Xing Yi regional grid in Guizhou Province, Southwest China. This plant was chosen for analysis as the plant is relatively new and performed well in the tests conducted in 2020, which showed that it was able to operate within both design and regulatory specifications [15,29].

2.2.1. Plant Design

The plant in Xing Yi is composed of 10 energy storage units of 2 MW/1 MWh each. These units operate in two parallel strings of five units each. Each five-unit module is composed of four 500 kW/250 kWh sub-modules. These modules are broken down into four clusters, with each cluster divided into seven battery modules. Each module is composed of 32 cells in series. The cell is the base unit, each capable of holding 92 Ah at a voltage of 3.2 volts and a design discharge rate of 2 C [30].

2.2.2. Performance of Nominal LFP Battery

The design and performance parameters from the Xing Yi plant are proprietary and not available through the open-source literature. Therefore, for the purposes of this analysis, performance data from an LFP battery is taken from a paper by Liao et al., who examined the electrochemical performance of LFP batteries and found that the specific capacity of such a battery was 122.1 mAh per gram at 3.2 V with a 0.2 C discharge rate. The researchers also tested such a battery at higher discharge currents and found that at a 10 C discharge rate, the battery reached 2450 charge/discharge cycles before reaching a 40% degradation of capacity [31]. A commercial LFP battery was also tested by Rehm et al., who found a lower specific capacity of commercially available LFP batteries, finding that a 42 g battery could hold 1800 mAh, or 42.86 mAh per gram, at 3.2 volts [32]. These laboratory test data allow for the calculation of the lithium content in the batteries.

2.2.3. Materials Required for Nominal Cell

The nominal battery cell, constructed by Liao et al., is composed of a carbon anode, an aqueous electrolyte composed of 1.3 M LiPF6 in a 1:3 mixture by mass, and a cathode composed of 80% LiFePO₄ [31]. The bulk of the weight of Li-ion batteries comes from the electrodes. The calculations performed assume that 75% of the weight is in the heavy LiFePO4 cathode, 15% of the weight is in the relatively light carbon anode, and the remaining 10% is in the aqueous electrolyte, yielding a value of 194.15 (g Li)/kWh. This translates into a total requirement of 1941.5 kg of Li for the nominal 10 MWh capacity energy storage plant.

The key metrics forming the basis for estimation of Li resource requirements and system cost are as follows:

Grid VRE capacities:

Reference case: Solar—18 TW; wind—11.90 TW.

High-growth case: Solar—21.25 TW; wind—14 TW.

Li requirement: 194 g/kWh.

This figure for (g Li)/kWh agrees well with results from Surendran et al. [33], who found that LFP batteries in a variety of cathode configurations and electrolyte types have a range of Li requirements, from 100 (g Li)/kWh to 230 g Li/kWh, with a mean of approximately 150 (g Li)/kWh across the various types. In addition, their team asserts that average Li usage in state-of-the-art battery packs ranges from 80 to 120 (g Li)/kWh. Therefore, the above figure for the lithium requirement is a conservative estimate for a well-designed system of LFP plants used for fixed energy storage.

3. Results and Discussion

3.1. Estimation of Energy Storage for Nominal Grid

Estimating the total amount of energy storage on a worldwide grid that is predominantly powered by solar and wind energy depends on many factors. These include the degree of interconnections between regions, the changes in weather patterns due to climatic effects, the dominant EV charging patterns, the design lifetime of battery plants, and the peak electrical demand, among many others. It is a complex challenge to arrive at a single figure for the amount of energy storage required for such a large grid. However, much work has been done analyzing smaller-scale grids. This analysis will take work from Xiao et al. from 2016 to attempt to estimate the approximate relationship between the required storage capacity (MWh) and the grid size (MW) [34]. Their work involves employing a modified IEEE 33-bus system to optimize the amount of storage needed from an economic perspective (i.e., whether it makes sense for a power company financially to build this amount of storage) [34]. Included in their analysis are multiple net present value calculations. Ultimately, the optimization algorithm developed by them suggests an energy storage requirement of 1332 kWh for 1900 kW of generation capacity comprising 58% solar and 42% wind power.

To estimate the total amount of storage capacity in the reference case, the above ratio is used to obtain the total of 22.23 TWh of storage required in the reference case. In the high-economic-growth case, the total grid size and amount of storage required is 25.69 TWh. This amount of storage corresponds to 2,223,000 10 MWh nominal plants required in the reference case and 2,570,000 nominal plants required in the high-economic-growth case.

Another crucial contributor to Li demand is the requirement to support transportation applications. Presently, this demand far exceeds the demand for fixed grid storage as the decarbonization of the transportation sector through transition to EVs leads that of other sectors. The IEA estimates that a total of approximately 1.45 million metric tons of Li will be required for EV applications in a net-zero-emission 2050 scenario [5]. This amount of Li must be taken into account when proposing an LFP-based grid storage system, as such an amount of Li will have a significant effect on the market.

3.2. Li Required to Support This Storage

To support the above amount of fixed storage, a total of 4.312 million metric tons of Li would be required in the reference case. For the high-economic-growth case, a total of 4.99 million metric tons would be required. These figures were arrived at by multiplying the required storage capacity postulated in Section 3.1 by the required amount of Li/kWh in Section 2.2.2. Vehicle storage would additionally require 1.45 million metric tons for a net-zero emissions by 2050 target.

3.3. Estimating Energy Required to Refine Li Resources

The current lithium market extracts lithium from its ores in a few ways. These include extraction through adsorption, electrochemical means, solvent extraction, and precipitation of lithium from brines [35]. Approximately 60% of the world’s lithium resources are in the form of lithium ions dissolved in surface or subsurface brines [35]. Alternatively, lithium can be extracted from lithium-yielding minerals, notably lepidolite and spodumene. Presently, a majority of lithium production is typically recovered from underground brine deposits. These liquid deposits contain a variety of alkali metals, including Li, and are pumped to the surface into evaporation ponds. The brine is allowed to sit in these ponds for a length of time, in the order of months, to allow the water content of the brine to evaporate. This evaporation uses only raw solar energy, akin to that used for recovering common salt. Once the brine has reached a high enough Li concentration, it is pumped to a processing facility for further refinement steps. These steps include filtration, ion exchange, further evaporation to concentrate the solution, and finally electropolishing to extract pure Li compounds [36]. These Li compounds are typically lithium carbonate or lithium hydroxide, and are shipped to battery manufacturers for their end use. Kelly et al. performed a lifecycle analysis of the energy required to process Li ore. It is estimated that 1 metric ton of LiOH · H₂O requires approximately 80,000 MJ of energy to process [37]. This implies that a total of around 580,000 GWh of energy would be required in the reference case, and around 670,000 GWh would be required with high economic growth. In 2019, the total world electrical demand was 22,848 TWh [38], making the reference case 2.54% of 2019 electricity demand and the high-economic-growth case 2.93% of 2019 demand. As world electricity demand has risen year on year, this percentage will shrink as time goes on. Such a figure, while a significant amount of energy, is only a small fraction of total output. However, such a large amount of energy would be expensive and likely require significantly uprated local sections of the power grid to supply such an intensely localized demand. Additionally, it is assumed that significant scale-up of lithium processing facilities will take place to keep up with future demand.

3.4. Analysis and Distribution of Li Resources and Extraction Upscaling Analysis

In 2023, the United States Geological Survey (USGS) estimated that world lithium reserves totaled 26 million metric tons of contained lithium [39]. These reserves are defined as sources that can be exploited economically with existing technology. The above calculation determining required lithium indicates that, to support continuous operation of a world grid composed primarily of wind and solar energy as well as account for EV storage, approximately 22.2% of the world’s lithium reserve would be required in the reference case, rising to approximately 24.8% for the high-economic-growth case. These numbers were arrived at by adding together the amount of Li required for stationary storage and that required for the transportation sector and dividing this by the 26 million metric tons of contained lithium as estimated by the USGS.

This amount is a significant fraction of current known lithium reserves. However, past experience has shown that as time passes, more resources are discovered and new technologies are developed that allow for exploitation of these resources. Therefore, the above fractions of known lithium resources will only decrease as 2050 approaches.

Michaux [40] asserts that the planet does not have enough Li resources to support both EV storage and backup of a renewable-based electric grid. However, using the postulated percentage of Li on a per-mass basis and the total postulated mass of batteries required for EVs and grid storage, his analysis yields a total Li requirement of only around 18 million metric tons.

Geographically, Li production is concentrated in a handful of countries. The most important area is South America (notably Bolivia, Argentina, and Chile), accounting for approximately 40% of global production [41]. Other important countries include the US (~25%), and Australia and China (~11% between the two) [41]. Smaller producers include Germany and the DRC. Although lithium resources are fairly concentrated, they are typically in countries that enjoy political stability, which translates into a small risk of supply disruption. Moreover, the two countries with the largest demand for lithium in the transition to green energy (China and the US), are able to supply a significant fraction of domestic consumption with domestic resources [42], removing significant amounts of supply chain risk. The major world producers of Li are shown in Figure 1.

The ability of the lithium production processes to expand significantly over a relatively short period (less than three decades) in order to satisfy such large increases in demand depends upon many factors. In some countries (for example, the US) all green energy projects, from dense housing to wind farms to transmission lines, face significant permitting and siting hurdles due to federal and state requirements for public comment and environmental reviews. This lengthens operational timelines significantly. Other countries, such as China and the South American countries with significant Li reserves, may not have such regulatory requirements, shortening the period needed for capacity expansion. The lithium market has so far avoided major shortages, with significant capacity added in recent years [42]. Experts predict shortages only in the 2030s, which is subject to future price signals and exploitation of still more available deposits. With a high enough market price, lower-grade reserves can be exploited profitably. Overall, the future of the lithium industry and its scalability depends on how much money decisionmakers are willing to invest in the green transition.

3.5. Implementation Costs

An LFP system has advantages over more conventional Li-ion batteries, namely that there is a reduced requirement for rare metals such as cobalt and manganese. Overall, this will be more cost effective. However, as with anything in engineering, there is a tradeoff associated with that gain. In this case, that tradeoff is slightly reduced performance as well as a less mature technology.

According to Ramasamy at the National Renewable Energy Laboratory, a standalone commercial Li-ion battery storage system is estimated to cost USD 806,000 for a 1200 kWh, 300 kW plant [43]. This figure includes engineering and development costs, sales tax, the battery cabinets themselves, installation costs, balance of plant costs, and inverter cost. This works out to USD 671.67 per kWh of storage. This figure is reasonable for an LFP plant, because even though the materials’ cost may be lower, the system does not develop as much voltage and hence requires more material to store the same amount of energy. Therefore, the rough overall cost for storage on a global grid scale in the reference case is around USD 15 trillion. In the case of high economic growth, the rough overall cost is USD 17.25 trillion. To put this number in perspective, Stockholm International Peace Research Institute (SIPRI) estimates global military expenditure in 2024 to be approximately USD 2.7 trillion, which is less than 2.5% of the world GDP [44]. It should be noted that costs for any process or system typically decrease with time as a result of technological advances along with the improvements in materials and manufacturing. Actual costs in future are likely to be lower than the limiting estimate stated above (https://data.worldbank.org/indicator/MS.MIL.XPND.GD.ZS, accessed on 11 July 2025).

In addition to the monetary cost, the system will incur environmental costs. The mining and refining of any mineral often require vast amounts of fresh water. A lifecycle analysis completed by Feng et al. determined the amount of freshwater required for refining 1 ton of lithium carbonate equivalent for various supply chains, stretching from ore mines in Australia to refining facilities in China. The average water consumption of the six supply routes investigated was approximately 166 cubic meters per ton of lithium carbonate equivalent [45]. This is an enormous amount of water for a comparatively small amount of lithium, and upscaling lithium ore production with these water consumption figures does pose an engineering and political challenge from a water availability and cleanliness standpoint.

3.6. Estimating Plant Lifetime

The degradation of Li-ion batteries over time is difficult to model on a macro level. Many factors, such as average depth of discharge, the rate of charging, the average level of charge, operating temperature, and the exact manufacture and geometry of the electrodes, impact the lifetime. With so many variables, plants that operate batteries typically have a manual detailing exactly how to operate the battery regarding state of charge, regular capacity tests, charging rate, and more, to attempt to optimize the lifetime of the battery.

Another factor regarding the lifetime of the nominal plant is when the engineers or plant operators decide the cells must be replaced. The International Standards Organization (ISO) defines battery end of life as 80% of the initial design capacity. However, in practice this number could be anywhere from 60 to 80% of the initial storage capacity, depending on how the economics works out for the plant stakeholders.

Omar et al. present a logical basis for approximating a battery’s cycle life. Assuming enough grid capacity for an average 50% depth of discharge (a balance between excess capital costs due to overcapacity and excess capital costs due to frequent replacement), they found that an LFP cell lasted for approximately 15,000 cycles until it was degraded to the ISO standard of 80% nominal capacity [46].

Determining how many nominal cycles such a plant would undergo per year is difficult to estimate on a large scale. The relative mix between wind and solar generation depends greatly on the regional weather patterns and the siting of the solar and wind plants, among many other factors. Weiꞵhar and Bessler [47] modeled an LFP microgrid and found that such a grid displayed significant day/night cycling, especially on summer days. Therefore, it seems reasonable to assume that the grid-level storage would also experience approximately one cycle per day. Such a plant would last 41 years. This lifetime is on par with other large industrial plants and even exceeds the design lifetimes of some fossil and renewable generators. It is anticipated that Li recycling technologies will achieve a level of maturity required to recycle most of the Li from batteries that have reached their end of life, similar to lead–acid batteries, where 99% of lead–acid batteries are recycled through safe and sustainable processes [48] (https://batterycouncil.org/battery-facts-and-applications/how-a-lead-battery-is-recycled/, accessed on 11 July 2025).

The above calculations were performed to determine if enough lithium physically exists to appropriately back up a renewable-dominated worldwide electric grid in 2050. The analysis accounted for transportation electrification and the lithium required for electric vehicle adoption. Additionally, the anticipated growth in energy demand was taken into consideration. The Li resource requirements presented in this analysis represent a limiting case based on current power generation efficiencies and capacity factors. Incorporation of an energy storage system will likely translate into higher capacity factors of the VREs, reducing the size of the world electricity grid and, consequently, Li requirements.

4. Conclusions

As Section 3 shows, the limiting factor for the considered system is not the lithium itself. Despite the enormous increases in power generation and lithium processing throughput required for global energy storage, there are no physical Li resource constraints on the implementation of a system consisting only of LFP batteries.

This conclusion is consistent with an analysis completed by Wentker et al. that looked into the criticality and risks associated with current battery technology and indicates that Li poses zero risk from the perspective of actual resource availability and proven reserves [49]. The calculated cost of a worldwide system seems substantial; however, if spread over two decades, the annual investment needed will be less than one-third of the global defense expenditure.

It should be noted that several factors will influence actual Li resource requirements, some of the significant ones being (1) improvements in Li battery technology reducing the specific Li requirement in a cell (g Li/kWh), (2) improvements in power generation efficiency resulting in a lower-capacity electric grid and consequent lower energy storage requirements than those used as a basis in this work, (3) deployment of other energy storage technologies (electrochemical, pumped hydro, hydrogen and other chemical, and so on) reducing the load on LFP batteries. However, in all these cases, the Li requirements will be less than those estimated in this study, reinforcing the inference that Li resources will not be a limiting factor. Li resources will not be a limiting factor unless the actual demand is more than four times greater than that forecast for the high-growth scenario considered in this work.

Overall, we have found that known Li resources that are amenable to commercial extraction using the present technology are adequate to meet sustainable energy storage goals based on a set of conservative assumptions surrounding LFP performance and storage plant design, future energy demand, storage requirements, and penetration of VRE sources.