Future Battery Material Demand Analysis Based on U.S. Department of Energy R&D Targets †

: The U.S. Department of Energy’s Vehicle Technologies Ofﬁce (VTO) supports research, development, and deployment of efﬁcient, sustainable transportation technologies that will improve energy efﬁciency and fuel economy, and enable America to use less petroleum. To accelerate the development and adoption of new technologies, VTO has developed speciﬁc targets for a wide range of powertrain components, including the energy storage system. In this study, we use Autonomie, Argonne National Laboratory’s (Argonne’s) vehicle system simulation tool to evaluate future energy storage requirements (power, energy, etc.) for different vehicle classes, powertrains, component technologies and timeframes. BatPac, Argonne’s tool dedicated to energy storage pack design and costs, is then used to quantify the materials required for each pack. Market penetrations are then used to estimate the overall material demand worldwide and in the United States, with or without recycling. The results demonstrate that the positive impact of VTO research and development will lead to signiﬁcant reduction in material compared to business-as-usual due to new anode and cathode designs, along with acceleration in battery cell chemistry penetrations. In terms of material demands, it is observed that lithium demand reaches about 80,000 tons (by a factor of 42–45), nickel demand reaches about 500,000 tons (by a factor of 47–56), manganese demand reaches about 30,000–50,000 tons (by a factor of 20–34), and cobalt demand reaches about 30,000 tons (by a factor of 13–28) in the future by 2050. The individual material demand per unit energy, however, decreases signiﬁcantly in the future due to advances in VTO research and development activities. The increase in battery material demands is mostly driven by increased electriﬁed vehicle ﬂeet penetration in the markets.


Introduction
Since the early 2000s, the yearly sales of plug-in hybrid electric vehicles (PHEVs) and battery electric vehicles (BEVs) in the United States have increased significantly. The year-to-year increase is on the order of a few magnitudes, and a significant share consists of BEVs [1]. Along with the increase in the PHEV market share observed in the United States, the market share held by PHEV sales has also increased in China and Europe [2]. As the number of electrified vehicle sales rises, the demands for lithium-ion cell batteries are also increasing. Between the cathode and anode of the battery cells, the percent share of lithium, nickel, manganese, and cobalt materials are the most significant in the cathode [3].
The U.S. Department of Energy's (DOE's) Vehicle Technologies Office (VTO) relies on different market penetration tools to evaluate material demands and supplies of different materials in the United States and globally. However, current methods of evaluating the market demand for battery materials do not apply individual VTO component targets, such as battery energy density, lightweighting, and so forth to estimate the overall impact of VTO research and development (R&D). Therefore, the evolution of current and future battery material demand relies on data that do not reflect the VTO technology target goals.
A large-scale study is being run as part of DOE's Benefits and Scenario Analysis (BaSce). Using the vehicles modeled as part of this study, further analysis has been conducted to estimate the battery material demand for different analysis years and the impact of recycling. It provides the range of estimates for business-as-usual (low technology progress) and VTO target goal impacts (high technology progress).
There are various studies that are ongoing to evaluate future material demands for lithium-ion batteries for the United States [4,5], as well as the global demand [6]. However, these models do not investigate the influence of battery chemistry evolution along with vehicle fleet penetrations. The study presented in this paper, however, presents an approach to quantify the impact of the DOE-VTO target goals for different technology advancements in evaluating potential future material demands from lithium-ion batteries. The study presented in this paper also takes into account different vehicle fleet penetrations to evaluate the total demand for both the U.S. market as well as at the global scale.

Large-Scale Simulation Process
The U.S. Department of Energy, Vehicle Technologies Office (U.S. DOE-VTO) generates the targets for advancements in technology and improvements in costs for engines, transmissions, batteries, fuel cell technologies, vehicle electrification, lightweighting, and so forth, over a given time frame. We use the different vehicle technology targets (battery energy density, battery energy specific cost, lightweighting, etc.) set by DOE-VTO to build the assumptions we evaluate over a range of time frames [7]. This paper will cover the results from 2015, 2020, 2025, 2030, and 2045 "laboratory years", which corresponds to the "model year minus 5 years". For example, a laboratory-year 2015 vehicle would reflect a vehicle that is available in the market in 2020, and similarly, a 2045 laboratory-year vehicle would be available in the market in 2050.
To implement uncertainties in the assumptions, two different set of targets have been implemented for all years: current/low technology progress (business as usual), and high technology progress (U.S. DOE-VTO targets).
The vehicle system simulation tool Autonomie [8] performs simulation on vehicle models that incorporate these vehicle technology targets. The vehicle models used for the simulation include power-split hybrid (split HEV), split plug-in hybrid (PHEVs), extendedrange electric vehicles (EREVs), and battery-electric vehicles (BEVs) of different all-electric ranges (AERs) in miles: BEV200, BEV300, and BEV400. These vehicle models are modeled across five different vehicle classes-compact, midsize, small sports utility vehicle (SUV), midsize SUV, and pickup. The Argonne Battery Performance and Cost (BatPaC) modeling tool [9] estimates the active material content of the different chemistries across the different analysis years. Figure 1 shows the detailed steps of the large-scale simulation process used to quantify the battery material demand using Autonomie and BatPaC.

Autonomie
Argonne's Autonomie tool is used to simulate the vehicles over the defined timeframe. The vehicles are sized for the given timeframe according to the component assumptions described earlier. A large-scale simulation approach is undertaken to evaluate the high volume of vehicle uncertainties. We use a distributed computing method that accelerates and facilitates the simulation runs [10].

BatPaC
BatPaC is a software modeling tool designed for policymakers and researchers who are interested in estimating the cost of lithium-ion batteries after they have reached a mature state of development and are being manufactured at high volumes. The tool captures the interplay between the design and the cost of these batteries for transportation applications.
BatPaC comes with a library of several lithium-ion battery chemistries and default inputs for all the parameters specified in different manufacturing areas of a factory.

Vehicle Sizing Process
Autonomie uses different vehicle sizing processes for different vehicle powertrains. For example, Figure 2 shows the detailed steps of the BEV sizing algorithm. We used the algorithm for the various BEV AERs (BEV 200/300/400 AERs). The main sizing algorithm for BEV sizing is as follows: • Battery and electric machine (EM) powers are sized to be able to follow the US06 cycle at low state-of-charge (beginning of charge sustaining mode) at 10% state-of-charge or to meet the requirement of acceleration performance. • The battery energy is sized to achieve the specified AER on the combined driving cycle, on the basis of adjusted energy values. • The vehicle weight is adjusted accordingly from updated electric-machine (EM) and battery weights, which are functions of the electric-machine peak power and battery energy.
The full vehicle sizing processes for the other electrified powertrains are specified in detail in the full report [10].

Battery Chemistry Assumptions
The battery cell chemistries analyzed in the study are:

Vehicle and Component Assumptions
This section details the different vehicle classifications and some of the major vehicle attribute selection used in the study. Table 2 details the different vehicle classifications defined for various performance times (0-60 mph time) in seconds, as well as corresponding vehicle attributes. The latest report from Argonne [10] details the assumptions and procedure involved behind the vehicle modeling and simulation efforts.  Table 3 summarizes the main DOE-VTO battery target assumptions associated with the different technologies over time. The different vehicles modeled in this study represent the laboratory years 2015, 2020, 2025, 2030, and 2045.  Table 4 details the assumption of different battery chemistry penetrations across the fleet for different analysis years.

Battery Material Recycling Assumptions
Some existing publications evaluate different battery recycling methods, with various and detailed modeling of the recycling processes. These studies detail different recycling methods, including hydro and pyro recycling methods [12,13]. In this study, we assume that 100% of the available battery energy is recyclable, assuming a 10-year end-of-life period for the batteries [14].  We can see that the battery pack weight decreases by almost 77% for split PHEV20s, 74% for EREV PHEV50s, and 65-70% for BEV 200/300/400 midsize vehicles. Vehicle weight is expected to decrease 26-42% across the different electrified vehicles. Due to advances in battery energy density, and other technology advances, we expect that the battery pack and the vehicle weights of the electrified vehicles will significantly decrease over time. With increasing battery energy density and vehicle lightweighting over time, the vehicle components results in smaller component sizes and weights and the compounding effects is observed in vehicle test weight reductions. Figure 6 shows the battery total energy for midsize vehicles across the different electrified powertrains We observed that the battery total energy requirement decreases significantly over time. For PHEVs, the battery pack total energy decreases by 33-50% for the 2045 laboratory year compared to the 2015 laboratory year. For BEV200s, the battery pack total energy decreases by 57% for 2045 laboratory year compared to the 2015 laboratory year. The decrease for the BEV400s approaches an 80% reduction in total battery pack energy. For high-range BEVs, the reduction is much greater due to the combined effects of advances in vehicle technology. The explanation for such observation is that, as different vehicle components improve over time, the different performance requirements (EV range, etc.) results in lower component sizes (battery energy, etc.). Figure 7 shows the impact of the sum of the total battery energy for the different electrified powertrains and vehicle classes modeled. The analysis is shown across the different laboratory years and technology progresses.  We can see that, because the advancement of technologies leading to reduced battery energy requirements over time and because of the advanced penetration of battery chemistries, the various material contents decrease over time across the different powertrains. Figure 9 illustrates the sum of different battery material contents across the different vehicle classes. It shows that similar effects are also observed when performing the analysis across vehicle classes.

Battery Material Contents
Using the assumed penetrations of the different battery cell chemistries across different laboratory years, when the sum of material contents is scaled to a total energy requirement of per GWh, Figure 10 shows the impact of VTO battery technology advances for different material demands per GWh.
It can be observed that the per-GWh contents of lithium, nickel, manganese, and cobalt contents decreases significantly over time, due to advances in battery technologies and penetration of advanced cell cathode chemistries.

United States Demand
Assuming there are 17 million new vehicle sales in the U.S. market every year [15], the appropriate market penetrations described in Chapter 3.3 have been adopted for the different vehicle powertrains and classes. Figure 11 illustrates the total battery energy requirement across the different years using our market penetration assumptions. The Figure shows that over time, we expect total battery energy requirements to increase significantly, due to the high penetration of electrified vehicles in the future. The current total battery requirements from this analysis align well with the status of today's market [16].  We observe that there is an increase in the various material contents by laboratory year 2045. This is driven by the increased penetration of electrified vehicles across laboratory years, even though the battery energy requirements decrease due to technology advancements. We further observe that with increasing AERs, the demand for material increases due to higher energy requirements. Figure 13 illustrates the total annual demand of different battery materials for U.S. across the vehicle classes modeled.
Across the vehicle classes, we see that market penetration has a significant influence on the material demands during different laboratory years. Figure 14 illustrates the total annual demand of different battery materials in the United States.
Our findings can be summarized as follows: •   Accounting for the recycling assumptions mentioned earlier, we further conducted analysis to evaluate the impact of recycling on the different battery material demands. Figure 15 illustrates the annual demand of different raw materials in the United States. to satisfy the market penetration of 17 million new vehicle sales per year, accounting for battery recycling. We see that the assumed recycling estimates reduce the material content demands significantly more with the advances from the aggressive DOE-VTO target.

Global Demand
If we assume that there will be 80 million new vehicle sales in the global market every year [17,18], we can adopt an appropriate market penetration for the different powertrains and classes. We evaluate the material demands at the global scale, scaling for the total energy requirements. Figure 16 illustrates the total annual demand for different battery materials around the world split up across the vehicle powertrains modeled.

Summary and Conclusions
In this study, we conducted an in-depth analysis of the demand for different raw materials for use in EV batteries. We show the impact that DOE-VTO goals for battery energy densities would have on reducing total energy demand, and the subsequent impacts on the demand for different raw materials. The impacts of the demand for different raw materials is influenced by the market penetration of vehicles, along with different battery chemistries. The main conclusions drawn from the study are outlined below: • The DOE-VTO goals are expected to decrease the total energy requirement of batteries by 31% by laboratory year 2045 for low technology progress and 40% for high technology progress; • The market penetration of electrified vehicles is assumed to increase by 62% by laboratory year 2045 for low technology progress and by 81% for high technology progress; • The DOE-VTO goals impact the total demand per GWh for different materials. By laboratory year 2045, total lithium content is expected to decrease by 16%, total nickel content by 62%, and total cobalt content by 75%; total nickel content is expected to increase by 5%; • Accounting for the market penetration of different battery chemistries and electrified vehicles, the total lithium content requirement is expected to increase by a factor of 42 by 2045 for low technology progress and 45 for high technology progress. The total nickel content requirement is expected to increase by a factor of 47 for low technology progress and 56 for high technology progress. The total manganese content requirement is expected to increase by a factor of 34 for low technology progress and 20 for high technology progress. The total cobalt content requirement is expected to increase by a factor of 28 for low technology progress and by 13 for high technology progress in both U.S. and global market.
For the purposes of this study, we made certain assumptions for the market penetration of vehicles, as well as the different battery chemistries. Although these assumptions could influence the values we obtained, this approach lets us evaluate the impact of DOE-VTO goals on the demand for different battery materials. Future research will include a closedloop recycling process integration to the study to enable accurate pyro and hydro recycling of the battery modules and the raw active materials.