Lithium-Ion Battery Environmental Impacts

Lithium-Ion Battery Environmental Impacts

21 Lithium-Ion Battery Environmental Impacts Linda L. Gaines1, *, Jennifer B. Dunn2 CE NTER FOR TRANSPORTAT ION RESEARCH, ARGONNE NATIONAL LABORATORY,...

2MB Sizes 2 Downloads 192 Views

21 Lithium-Ion Battery Environmental Impacts Linda L. Gaines1, *, Jennifer B. Dunn2 CE NTER FOR TRANSPORTAT ION RESEARCH, ARGONNE NATIONAL LABORATORY, ARGONNE, IL, USA, 2 SYSTEMS A SSESSME NT SECTION, ARGONNE NAT I ONAL LABO RAT ORY , ARGONNE, I L, USA * CORRESPONDI NG AUTHOR: [email protected] . GO V 1

CHAPTER OUTLINE 1. Introduction ................................................................................................................................... 483 2. Benefits of Lithium-Ion Battery Recycling .................................................................................. 484 3. Environmental Impacts of Lithium-Ion Batteries ....................................................................... 486 3.1. Battery Composition.............................................................................................................. 486 3.2. Battery Materials’ Supply Chain ........................................................................................... 487 3.3. Battery Assembly ................................................................................................................... 490 3.4. Contribution of Battery to Electric Vehicle Life-Cycle Environmental Impacts ............... 493 4. Overview and Analysis of Lithium-Ion Battery Recycling Technologies ................................. 495 4.1. Pyrometallurgical Recycling Process..................................................................................... 496 4.2. BIT Recycling Process ............................................................................................................. 497 4.3. Intermediate Physical Recycling Process .............................................................................. 498 4.4. Direct Physical Recycling Process .......................................................................................... 499 4.5. Analysis of Recycling Processes............................................................................................. 500 5. Factors that Affect Recycling ....................................................................................................... 504 6. Conclusions .................................................................................................................................... 506 Acknowledgments ............................................................................................................................. 506 Nomenclature ..................................................................................................................................... 507 References........................................................................................................................................... 507

1. Introduction Battery-powered vehicles are promoted by many countries, including the United States, China and European countries, as a means to foster energy independence and, potentially, reduce the greenhouse gas (GHG) emissions stemming from the transportation sector. Lithium-Ion Batteries: Advances and Applications. http://dx.doi.org/10.1016/B978-0-444-59513-3.00021-2 Ó 2014 Elsevier B.V. All rights reserved. The contribution has been prepared by UChicago Argonne, LLC, Operator of Argonne National Laboratories under contract No. DE-AC02-06CH11357, with the U.S Department of Energy.

483

484

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

The possible advent of large-scale growth in battery-powered vehicle deployment provides a unique opportunity to delve into the battery supply chain and examine it for any environmental issues or other roadblocks pertaining to material scarcity that may arise as barriers. One key concern is that batteries incorporate materials whose production may incur adverse environmental impacts and consume large amounts of energy. Additionally, if demand for these materials (especially lithium or cobalt) for batteries outpaces supply, scarcity may become an issue and impact the price and feasibility of large-scale adoption of battery-powered vehicles. Lithium-ion (Li-ion) battery recycling could be one way to relieve supply constraints and mitigate the environmental impact of virgin material production. It may be possible to recover other materials in addition to cathode materials, such as the anode, the electrolyte and structural materials such as aluminum, steel and plastics from the battery during recycling. Recycling is one of the canonical 3 R’s—reduce, reuse, recycle—but it’s not always the best option for used items [1]. So we need to identify the benefits of recycling for any given product before just assuming that it’s the “greenest” thing to do. For Li-ion batteries, recycling offers several potential benefits. These are discussed in general terms here, but we will demonstrate later that the actual benefits depend on several factors, most notably battery chemistry and the recycling process chosen. In this chapter, we consider how recycling lithium from batteries could impact lithium demand. Additionally, we examine the cradle-to-gate (CTG) environmental impact of Li-ion battery production. Finally, we estimate the impact of different Li-ion battery recycling methods on the CTG environmental impacts of Li-ion battery production to evaluate the actual benefits available from recycling. We also consider factors that could enable or hinder efforts to make optimal Li-ion battery recycling the norm worldwide.

2. Benefits of Lithium-Ion Battery Recycling As described in the introduction, growth in Li-ion battery deployment in the transportation sector could strain supplies of lithium and other metals. The supply of materials on Earth is finite, and reusing something that has already been extracted reduces the demand for virgin material and delays the day when easily accessible supplies are exhausted. Figure 21.1 [2] shows potential demand out to 2050 for lithium for automotive batteries in the United States, under a scenario in which vehicles with electric drive achieve unprecedented success, representing 90% of new car purchases (red line). This scenario represents an upper bound on the potential US demand. The figure also shows how much recycled material would be available if all of the battery materials were recovered after 10 years (gold line), when the batteries were no longer expected to be suitable for automotive use, or 5 years later (blue line), after a second use such as utility storage. Subtracting the recovered material from the virgin material provides the net demand for virgin material when recovered material is substituted for it (dotted lines). The quantity of virgin material needed not only peaks at a lower value, but even turns

Chapter 21 • Lithium-Ion Battery Environmental Impacts

485

60,000 US battery demand

50,000

Available for recycling (10 years) Net virgin material demand (10 years)

Tonnes lithium

Available for recycling (15 years)

40,000

Net virgin material demand (15 years)

30,000

20,000

10,000

0 2010

2015

2020

2025

2030

2035

2040

2045

2050

FIGURE 21.1 Maximum potential US lithium demand and supply from recycling. (For color version, refer to the plate section.)

down when growth of demand slows. Note that if the material is reused, the net virgin material peak occurs later, and is higher, than if the material is recycled immediately. This example is for lithium, but the same general behavior would be true of any recyclable material. Lithium may not be scarce, but cobalt supplies are declining, and copper and possibly even aluminum [2,3] could become scarce as well without recycling. So recycling can help to mitigate our ever-growing need for materials. There is also a benefit from recycling in terms of national security. For many materials, including lithium, the US relies on imports for much of its supply, as do many other countries. In-country recycling can reduce dependence on imports. In many cases, recycled material is less costly to produce than virgin material, and its use could be one way of reducing high battery material costs. This is true of manufacturers’ home scrap as well as postconsumer material. There is an obvious caveat here: the recycled material must achieve acceptable performance specifications, and battery manufacturers, for whom product quality is paramount, may be reluctant to use recycled material even if it’s cheaper. Product quality is therefore a key concern for emerging recycling companies, whose new-business creation represents another potential economic benefit from recycling. Reduction of waste disposal costs (both manufacturer and end-of-life) is yet another. Recycling, especially in Europe, is often mandated by law, motivated by perceived environmental benefits [4]. Most of the information provided in this chapter is devoted to evaluating the actual energy use and emissions benefits that could be achieved by recycling of automotive Li-ion batteries. To do this, we first examine primary production processes in the battery supply chain (including mining, transport, raw material processing and manufacturing) and then several processes for recycling back to battery-grade material, for comparison, in order to see how benefits can be maximized.

486

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

3. Environmental Impacts of Lithium-Ion Batteries In this section, following the framework of life-cycle analysis [5–7], we develop our process-level analysis of the energy and environmental impacts of producing automotive Li-ion batteries from new materials. Material and energy flows were calculated for each step in the battery supply chain, as described in detail elsewhere [8]. Figure 21.2 represents the flows required to characterize each of the many processes in the supply chain. These flows were combined in Argonne National Laboratory’s Greenhouse Gases, Regulated Emissions and Energy Use in Transportation (GREET) model. From the energy consumed on-site at mining and manufacturing facilities and during transportation of raw materials and products, GREET calculates the full fuel-cycle energy, which includes, for example, energy used in the production of coal. Similarly, for materials consumed, GREET rolls the upstream energy and environmental impacts of their production into the analysis. The focus of this assessment is on energy consumption and emissions to air, including GHGs.

3.1.

Battery Composition

The ingredients of a Li-ion battery depend on the desired performance characteristics. For a high-energy application such as a battery for a battery electric vehicle (BEV) without an internal combustion engine, specific energy is the key property to maximize. Alternatively, for a plug-in hybrid electric vehicle (PHEV), the battery specific power is more critical. The battery performance characteristics will determine the amount of active materials and the requisite structural materials that contain them within the battery. We used Argonne National Laboratory’s Battery Performance and Cost (BatPaC) model [9,10] to develop a materials inventory for hybrid electric vehicle (HEV), PHEV and BEV batteries. BatPaC represents the present-day technology and assembly practices, with some efficiency improvements that yield a more energy-dense battery assumed suitable for market-scale production. In BatPaC, an aluminum foil is the current collector at the cathode, whereas the anode current collector is a copper (or in some cases aluminum)

FIGURE 21.2 Schematic of stages, inputs and outputs of a life-cycle analysis. (For color version of this figure, the reader is referred to the online version of this book.)

Chapter 21 • Lithium-Ion Battery Environmental Impacts

487

foil. BatPaC allows users to select from several cathode materials: lithium manganese oxide, lithium iron phosphate, lithium nickel manganese cobalt oxide or lithium nickel cobalt aluminum oxide. Graphite coats both sides of the anode. A polymeric material (polyvinylidene fluoride [PVDF]) binds together the active-material particles. The battery assembly process uses N-methyl-2-pyrrolidone (NMP) as a solvent to facilitate contact between the PVDF and the active materials. A porous polymer membrane separates the two electrodes. An electrolyte composed of LiPF6 and the solvents ethylene carbonate (EC) and dimethyl carbonate (DMC) fills the pores in the separator and the active materials. A pouch made from polyethylene terephthalate, aluminum and polypropylene encloses the cells. Multiple cells (16 in our model) are combined into a module housed in aluminum. All of the battery modules are contained in an aluminum/insulation jacket. Compression plates and straps are steel. We used this BatPaC design to construct a material inventory for a battery containing all of the above-described parts and a battery management system. BatPaC inputs and the resulting battery compositions are given in Table 21.1 and Table 21.2, respectively.

3.2.

Battery Materials’ Supply Chain

For a full picture of the environmental impact of making Li-ion batteries, it is essential to consider each step in the supply chain, going back to the source of lithium. Figure 21.3 displays the processes and system boundary included in the assessment of CTG environmental impacts of battery production. The battery supply chain is spread around the world, with much battery production occurring in Asia. We conducted our analysis, however, in a US context, with battery assembly occurring in western Michigan, which is home to several battery manufacturers (e.g. LG Chem and JCI). We assume for our base case that the cathode active material is made domestically in the US from firing of Li2CO3 and Mn3O4. Other metal oxides will be discussed briefly. The US imports more than 50% of its Li2CO3 [11]; 47% of these imports are from Chile. We have therefore considered two possible sources of Li2CO3. The first is from Salar de Atacama, Chile. The alternative source is domestic production of Li2CO3 in Nevada. Figure 21.3 includes the production of the cathode material from lithium brine in Chile by SQM, the world’s largest lithium producer. In Chile, the brine extraction process

Table 21.1

Battery Parameters Modeled in this Study

Power (kW) Energy (kWh) Mass (kg) Specific power (W/kg) Specific energy (kWh/kg) Range (km)

HEV

PHEV

BEV

30 2 19 1500 0.10 N/A

150 9 89 1715 0.11 48

160 28 210 762 0.13 160

488

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

Table 21.2

Material Inventories for HEV, PHEV and BEV Batteries

Component

HEV

Percent Mass PHEV

BEV

Lithium manganese oxide (LiMn2O4) Graphite/Carbon Binder Copper Wrought aluminum Lithium pentafluorophosphate (LiPF6) Ethylene carbonate (EC) Dimethyl carbonate (DMC) Polypropylene Polyethylene Polyethylene terephthalate Steel Thermal insulation Glycol Electronic parts Total battery mass (lb)

27%

27%

33%

12% 2.1% 13% 24% 1.5%

12% 2.0% 15% 22% 1.6%

15% 2.5% 11% 19% 1.8%

4.4% 4.4% 2.0% 0.26% 2.2% 2.8% 0.43% 2.3% 1.5% 41

4.7% 4.7% 2.2% 0.40% 1.6% 1.8% 0.33% 1.2% 0.9% 196

5.3% 5.3% 1.7% 0.29% 1.2% 1.4% 0.34% 1.0% 1.1% 463

uses mostly diesel-powered pumps to pump aqueous solutions of brine into a series of evaporation ponds [8]. Produced in tandem with lithium brine are potassium chloride, potassium sulfate and boric acid. We allocate the burdens of diesel consumption at SQM among these coproducts on a mass basis. The lithium brine is transported to Antofagasta, Chile, where it is converted to Li2CO3 through a series of extraction, precipitation and filtration steps. Figure 21.3 displays the process chemicals that are consumed in the production of Li2CO3. Soda ash, which is consumed at a level of 2.48 kg/kg Li2CO3, is transported to Chile by ship from the western United States, where Wyoming is a major producer of this compound. Li2CO3 production in Nevada from brine is similar to that in Chile. The data we obtained for Nevada-based production, however, indicated that higher quantities of residual oil and soda ash are consumed in Nevada than in Chile. Lithium brine can also be sourced from spodumene, a mineral consisting of lithium aluminum silicate or LiAl(SiO3)2. Production from minerals was abandoned when less costly production from salars was introduced. However, production from minerals is again a possibility and would increase and diversify supply and increase the domestic US production. One reason spodumene-derived lithium is more expensive is that it is more energy-intensive to produce. In addition to mining and milling the ore, the mineral spodumene must be treated at 1000  C to achieve the structural transformation from an alpha to a beta form to enable acid leaching using sulfuric acid.

Chapter 21 • Lithium-Ion Battery Environmental Impacts

Anode active material

Cathode active material

Binder

Electrolyte

489

BMS

Soda ash Lithium brine

Lime

Lithium carbonate

HCl

Mn2O3

Pet coke

PVDF (binder)

Graphite

NMP (binder solvent)

LiMn2O4

H2SO4

LiPF6

BMS

Ethylene carbonate Dimethyl carbonate

Material production

Alcohol

Assembly

Use Aluminum Steel

Recycling/re-use/ disposal Materials production

Copper

Pyrometallurgical

Battery assembly Battery recycling

Thermal insulation Plastics

Hydrometallurgical Intermediate physical

Battery use (not included)

Direct physical

FIGURE 21.3 System boundary for the cradle-to-gate environmental impact analysis of lithium-ion batteries, which includes the materials production and battery assembly stages. (For color version of this figure, the reader is referred to the online version of this book.)

Lithium is then recovered in the form of lithium salts. This analysis does not include spodumene-sourced lithium. Figure 21.4 compares the total energy consumption in the production pathways of LiMn2O4 using Li2CO3 from Chile and from Nevada, taking into account all transport of process ingredients and final products. Energy consumed in transportation is 14% and 6% of total energy consumption in the LiMn2O4 supply chain with Chilean- and US-derived Li2CO3, respectively. One major contributor to transportation burdens in the Chilean material supply chain is the soda ash transport from the US to Chile. The steps for production of Mn2O3 (by heating manganese ore in a kiln) and LiMn2O4 are identical for the two routes to the final active material. LiMn2O4 is approximately 16% more energy-intensive to produce with US-derived Li2CO3. This additional energy burden is not very large, and given that domestic energy security is enhanced through use of domestic materials, US-based Li2CO3 production is feasible. The anode material considered, graphite, is made from a high-temperature process with petroleum coke and hard pitch as feedstocks. Carbonaceous anodes, however, can have different forms in Li-ion batteries. Natural graphite, hard carbon, soft carbon and mesocarbon microbeads, using fossil fuels are widely used for lithium insertion anodes in commercial cells. All synthetic graphite materials require 2700  C for full graphitization, so this fossil fuel-based process is energy-intensive. Recently, coating with very thin layers of

490

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

FIGURE 21.4 Contributions to energy intensity of LiMn2O4 production with Li2CO3 from Chile and Nevada. (For color version of this figure, the reader is referred to the online version of this book.)

amorphous carbon has emerged as a viable way to protect the surface of carbonaceous anodes against deterioration under cell working conditions. This process uses gas-phase sources, such as propylene or methane, which need to be cracked at 700  C in the presence of graphite. Lithium titanate (Li4Ti5O12) material has recently garnered attention as a high-power alternative anode material in Li-ion cells, where energy is less of an issue. It is produced by reacting lithium carbonate (Li2CO3) and titania (TiO2 in its anatase crystalline form) at 850  C under air. This process may require less energy than graphite production. We based our calculations of the energy and emissions intensity of graphite production on data for the aluminum industry, which uses graphite electrodes comparable to those used as Li-ion battery anodes. Another key battery component is the electrolyte. While many possible electrolytes are being developed, we have selected lithium hexafluorophosphate (LiPF6) in a solvent of EC and DMC. Little information exists on the production of LiPF6, so its impact, although estimated on the basis of data presented by Espinosa et al. [12], is somewhat uncertain. Given the potentially harmful nature of the compound, it is of interest to consider what its fate may be during the battery recycling process, which we address in Section 4.

3.3.

Battery Assembly

The battery assembly process is the second step of the CTG portion of the battery’s life cycle and is described in detail by Nelson et al. [10]. The first step in battery assembly is the preparation of electrode materials by mixing. The electrodes are then coated with the active material using the binder (PVDF) and binder solvent (NMP). In the next step, the coated electrode passes through a furnace where the NMP evaporates; 99.5% of it is recovered and reused [10]. The coated electrodes are calendered and slit. After a vacuum

Chapter 21 • Lithium-Ion Battery Environmental Impacts

491

drying step, they enter a dry room, where cells are stacked, current collectors are welded, and the cells are enclosed in containers. Next, the electrolyte is injected and the cell is closed. The cells undergo formation cycling to assess their performance; poorly performing cells (about 5% of the total) are rejected. The cells are sealed and undergo charge retention testing prior to being assembled into modules. The battery packs are assembled and tested as the final step in the process. In examining this process, we considered the dry room to likely be the main energy consumer in an assembly plant. We therefore estimated dry room energy consumption on the basis of a quote we obtained from SCS Systems for a dry room [8]. We also calculated the energy required for formation cycling of the cells and assumed that the energy consumed to operate the dry room and the formation cycling equipment is 60% of the total assembly plant energy consumption. Results are not greatly affected by this assumption, as assessed in a sensitivity analysis [13]. On the basis of our CTG analysis of the production of Li-ion batteries with LiMn2O4 cathodes on a per-mass-of-battery basis, we determined that structural materials (aluminum and copper) contributed about half of the energy consumed in the battery production chain, as shown in Figure 21.5(a). The cathode material production consumed roughly 10–14% of CTG energy. The contributions of the electrolyte (LiPF6, EC, DMC) ranged from 9% to 13%. The battery assembly process contributed about 6% of the CTG energy, confirming that reducing the energy intensity of battery components, potentially through using recycled materials, is a promising route to lowering battery CTG energy. Results are similar for CTG GHG emissions (Figure 21.5(b)). A question under current consideration is how the production energy intensity of the Li-ion battery cathode will change depending on the active materials chosen. It will be important to consider leading and potential future cathode materials including LiCoO2, LiFePO4 and lithium–nickel–manganese–cobalt chemistries in further analyses. Many available cathode materials are made by calcination of mixtures of lithium carbonate and transition-metal precursors at high temperatures (600–800  C). Lithium hydroxide is also used, with special safety procedures necessary during the process of mixing. The choice of cathode material will impact not only energy and emissions associated with the CTG steps of battery production, but also the benefits derived from battery recycling, as described in Section 4. Table 21.3 shows the energy density of three cathode materials and estimates of the energy to produce them. Although LiCoO2 is more energy-intensive to produce as a result of the large energy intensity of cobalt oxide production [14], it may be required in lower amounts in Li-ion batteries owing to its higher energy density. LiFePO4 is more energy-dense than LiMn2O4 and could be less energy-intensive to produce. Batteries using this cathode material may have a lower CTG energy intensity than batteries with LiMn2O4 cathodes. However, we will see that production of the cathode is a relatively small component of the entire vehicle life-cycle energy use and of (most) emissions, and so will not significantly influence final results.

492

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

(a)

(b)

FIGURE 21.5 (a) Life-cycle energy consumption of battery components (MJ/kg battery), and (b) Life-cycle GHG emissions of battery components (g CO2e/kg battery). Components marked with an “a” make up the electrolyte. Polyvinylidene fluoride and N-methyl-2-pyrrolidone (NMP) are the binder and binder solvent, respectively. (For color version of this figure, the reader is referred to the online version of this book.)

Chapter 21 • Lithium-Ion Battery Environmental Impacts

493

Table 21.3 Cathode Material Energy Density and Energy Intensity of Production

3.4.

Cathode Material

Energy Density (mAh/g) [9]

Estimated Energy Intensity of Production (MJ/kg)

LiMn2O4 LiFePO4 LiCoO2

100 150 150

30 [8] 20 [15] 150 [13]

Contribution of Battery to Electric Vehicle Life-Cycle Environmental Impacts

In Figure 21.6, we place the CTG energy consumption of the Li-ion battery in the context of the fuel and vehicle cycles of a PHEV and a BEV. Assessing the battery contribution in this context is important to avoid adverse environmental impacts as the market share of

FIGURE 21.6 Total energy consumption (a), GHG emissions (b), and SOx emissions (c) for the vehicle and fuel cycles of PHEV and BEVs powered by the US and California grids. (For color version, refer to the plate section.)

494

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

Table 21.4

Residual oil Natural gas Coal Nuclear power Biomass Others

Default GREET Electricity Grid mix for US and California [14] US Electric Grid

California Electric Grid

1.0% 22.9% 46.4% 20.3% 0.2% 9.2%

0.0% 41.0% 8.1% 23.1% 0.9% 26.8%

battery-powered vehicles increases. In our calculations, the vehicle lifetime is 260,000 km with no battery replacement. (Battery replacement would double impacts.) Analyses were conducted with the default GREET values for the compositions of the US and California grids, shown in Table 21.4. The California grid is less carbon-intensive. In terms of energy use and GHG emissions, the contribution of the battery (about 3%) is greatest in the full life cycle of a BEV powered from the California grid. The battery contribution is just under 1% of the total fuel and vehicle cycle energy consumption and GHG emissions for a PHEV. The 8% contribution of the battery to life-cycle SOx emissions, though, is notable for a BEV powered by the California grid. The key contributors to CTG SOx emissions in the battery life cycle are copper and aluminum (Figure 21.7). Two routes are possible to reduce the CTG energy consumption impacts of Li-ion batteries: first, provide lower-energy-intensity raw materials, either through improving the energy efficiency of the existing supply chain or by substituting less energy-intensive

FIGURE 21.7 CTG air emissions attributable to key lithium-ion battery components. (For color version of this figure, the reader is referred to the online version of this book.)

Chapter 21 • Lithium-Ion Battery Environmental Impacts

495

materials. A second possible route is battery recycling. Until recently, estimates of the energy intensity of battery components produced through battery recycling were scarce in the literature. For three of the recycling processes described in the following section, we have developed estimates of the energy intensity of recovering cathode and structural materials from Li-ion batteries with LiMn2O4 as a cathode.

4. Overview and Analysis of Lithium-Ion Battery Recycling Technologies Recycling can recover materials at different production stages, from basic building blocks to battery-grade materials. The different recycling processes differ in how many of the primary material production processes can be avoided when new batteries are made from used ones, as can be seen in Figure 21.8. In general, the recycling processes are not

Natural gas and petroleum

Ores

Li2CO3 manufacture

Organic solvents

(Al) SO2

Petrochemical manufacture

Polymers

Lithium brines

Na2CO3

(Solvents)

Lithium salts

Primary metal production

Li2CO3

Co, Ni, Mn, etc.

Al, Cu

Anode carbon Fabrication

Electrolyte production

Cathode material manufacture

Separator

Electrolyte

Cathode material

Rolling

Current collectors

Coating winding

Electrode

Smelting Assembly, testing

Intermediate process Direct recovery

Finished cells

FIGURE 21.8 Schematic flow chart for the production of lithium-ion cell materials, where purple ovals and light blue rectangles represent component materials and process steps, respectively. The red, yellow, and green symbols next to various components indicate where new materials can be replaced by smelting, by the intermediate process, and by direct recovery, respectively, and the corresponding shaded outlines encompass the process steps that are avoided by each of these alternative flows. (For color version, refer to the plate section.)

496

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

resource-intensive, so the more primary production steps can be avoided, the lower the impacts associated with the recovered products. These recycling processes are also at different stages of development. At one extreme are commercial pyrometallurgical processes (smelting) that recover basic elements or salts. At the other extreme, direct physical recycling that can recover battery-grade material has been demonstrated at bench scale. The quantitative analysis includes three recycling processes: a process under development at the Beijing Institute of Technology (BIT) [16] and intermediate and direct physical processes. We use the term intermediate to indicate that cathodic active material can be obtained from upgrading of process outputs with simple chemical processes, and direct to indicate that process outputs can be reincorporated into batteries with little or no additional processing. The materials recovered from these processes could be incorporated into the broader economy (open-loop) or reused in Li-ion batteries (closed-loop). Currently, we consider the latter. Note that none of the processes analyzed was specifically designed to handle batteries with LiMn2O4 chemistry. In our current work, we assume these technologies could recycle batteries with LiMn2O4 cathodes and examine the role they might play in reducing energy and emissions associated with battery material production and assembly. As we expand our analysis to include additional cathode materials, we will reexamine these recycling processes. Next, we describe the types of battery recycling processes that have been considered.

4.1.

Pyrometallurgical Recycling Process

Smelting is currently operational on a large scale and can take just about any input, including different battery chemistries (Li-ion, nickel metal hydride etc.) or mixed feed. Figure 21.9 is a simplified flow chart of this commercial process [17]. Umicore, a European company, collects spent batteries and feeds them into its high-temperature smelter in Hoboken, Belgium, with no preprocessing except removal from the outer pack. Organic components in the batteries (plastics, electrolyte solvents and carbon anodes) are burned; the heat fuels the smelter and the carbon serves as a reducing agent for some of the metal. The main products (from current cathode materials) are cobalt and nickel, which can then be sent to a refinery in Olen, Belgium, where the CoCl2 is made; it is subsequently forwarded to South Korea to produce LiCoO2 for batteries (using new purchased lithium). The recovered metals are also suitable for other uses. Their recovery not only saves about 70% of the energy needed for their primary production from sulfide ores, but also avoids the significant SO2 emissions from such production. It is also the economic driver for the process. Other metals, such as copper and iron, can be recovered as well. The lithium and aluminum from the smelter currently go to the slag, now used as an additive in concrete, but the company is investigating hydrometallurgy to recover lithium for higher-value uses such as batteries. However, recovery from slag may incur higher costs and energy use than production from brines. Waste gases are subjected to

Chapter 21 • Lithium-Ion Battery Environmental Impacts

497

Stack gases Gas cleaning

Dismantled lithium-ion batteries

Air Limestone Sand Slag

Slag including Li Smelter 1400 °C

Potential recovery for use in cement HCl

Alloy: Co, Cu, Ni, Fe

Leaching 1

Leaching 2

Natural gas

Cu

Solvent CoCl 2 extraction

Fe

Oxidation

Oxidizing agent

Co3O4

Ni to Ni(OH)2

Li2CO3

Firing 900 °C

LiCoO2

Natural gas

FIGURE 21.9 Pyrometallurgical recycling process flow sheet. (For color version of this figure, the reader is referred to the online version of this book.)

high temperature to avoid emissions of dangerous organics like furans or dioxin. The fluorine in the electrolyte and binder is incorporated into inert compounds. The company claims a 93% recovery rate for Li-ion batteries (metals 69%, carbon 10%, plastics 15%), but a much smaller percentage actually comes out as usable high-value material. This process is not examined quantitatively in this chapter because it is not economical to recycle batteries with LiMn2O4 cathode materials.

4.2.

BIT Recycling Process

In the BIT process, which is currently under development [16], battery components are separated through a combination of mechanical, high-temperature and chemical steps, as shown in Figure 21.10. First, spent battery cells are discharged and, at the bench scale, manually separated into anode and cathode components. The aluminum is separated from the cathode material after soaking in a warm NMP bath. The cathode is then calcined, the product is ground, and the metals are leached in a hydrometallurgical step with hydrogen peroxide (to enhance metal solubility) and an organic acid. Acid and oxidant concentrations are not optimized, as they would be in an industrial process. We assume that 90% of the acid can be recovered and reused. We also assume that the lithium recovered in this manner can be separated from cobalt, precipitated as Li2CO3, and then fired with Mn2O3 to produce LiMn2O4. It is not certain what the fate of fluorinated compounds would be in this process. We reiterate that this process is currently at the bench scale and its analysis should be considered preliminary. Additionally, manganese recovery from this process is undemonstrated, so we assume manganese is not recovered.

498

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

FIGURE 21.10 BIT recycling process flow sheet. (For color version of this figure, the reader is referred to the online version of this book.)

4.3.

Intermediate Physical Recycling Process

This approach to battery recycling is by its nature a low-energy process. Very little, if any, heat is employed; the energy that is required is used for running shredders and pumps and for producing liquid nitrogen. An example of this approach, the Toxco recycling system, is shown in Figure 21.11 [18]. This process is in commercial operation in Trail, B.C., and additional capacity is under construction in Lancaster, Ohio. Toxco’s approach involves separating batteries by chemistry, shredding them (reducing their size) in a flooded hammer mill, and separating product streams through a combination of a shaker table and two filters. Depending on the feed, liquid nitrogen is used only in two special circumstances: (1) to suppress potential ignition of electrolyte in discarded batteries with high residual energy (such as Li metal) and (2) to achieve better materials separation when certain battery potting materials were used.

Chapter 21 • Lithium-Ion Battery Environmental Impacts

Scrubber and filter

499

Air emissions

Spent batteries Shredder

Hammermill

Filter tank Li brine Evaporator filtrate and storage tank array

Anode, cathode material

Shaker table

Mixed Cu/Co plastics Al

Carbon filter press

Mixed metal oxides and carbon

Soda ash Mixing tank

Filter press

Waste water

Li2CO3 FIGURE 21.11 Intermediate recycling process flow sheet. (For color version of this figure, the reader is referred to the online version of this book.)

The Toxco system produces three product streams: (1) mixed materials comprised of steel, paper and plastics; (2) intermediate materials made up of collector foils, mixed metallics and a little electrode material; and finally (3) a slurry of cathode and anode materials, some of which is carbon. Streams (2) and (3) are of the most value. Substantial amounts of copper, cobalt, nickel and aluminum are contained in stream (2). Stream (3) produces two fractions: one contains primarily a cathode material and anode carbon mix, which is rich in cobalt (35% by weight), and the second is a filtrate, from which lithium carbonate (Li2CO3) is recovered. This material is currently used for other products, but could be purified for use in batteries. The fate of solvents and fluorinated compounds in this process is uncertain.

4.4.

Direct Physical Recycling Process

This approach to battery recycling is a low-temperature process with a low-energy requirement. The components are separated by a variety of physical and chemical processes, and all active materials and metals can be recovered. Only the separator is unlikely to be usable, because its form cannot be retained.

500

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

Spent batteries

Discharging and disassembly

Supercritical CO2 Surfactants and water

Spent cells

Extraction

Size reduction

Physical separation

Al, Cu, Fe

Cathode materials

Anode carbon Plastics

Flash Recycling Li2CO3

Electrolyte organics

Li-ion battery manufacturing

LiMn2O4

Relithiation

FIGURE 21.12 Direct recycling process flow sheet. (For color version of this figure, the reader is referred to the online version of this book.)

The first step in this process, illustrated in Figure 21.12, is the discharging of batteries and their disassembly to the cell level. Next, CO2 is added to a container that contains breached, discharged cells. The temperature and pressure are raised to bring CO2 above its critical point (31  C, 72.8 atm). The supercritical CO2 extracts the electrolyte (EC, DMC, LiPF6) from the cells before being vented to a different chamber. The temperature and pressure are reduced and the electrolytic compounds separate from the gaseous CO2. It has been demonstrated that after further processing, these can be recycled and used again in batteries. Subsequently, the cells are subjected to a number of mechanical steps, possibly in the absence of water or oxygen, which would break the cathode compounds apart. Next, techniques that exploit differences in electronic conductivity, density or other properties are used to separate the cell components. It is unclear how the PVDF binder separation from the active materials, which may prove to be a significant barrier for this process, occurs. Cathode materials can be reused in batteries after some relithiation. Careful segregation by chemistry of the process feed will be required to insure product purity and value.

4.5.

Analysis of Recycling Processes

There are some key differences among recycling processes that will impact their effectiveness, flexibility in the face of changing battery technology, and overall likelihood of success. Table 21.5 summarizes these differences. Pyrometallurgical processing produces products that are suitable for any market, and is the only recycling method that can easily accommodate mixed battery chemistries, but it depends on cobalt recovery to be economical, and leaves the lithium in slag, from which its recovery would be expensive

Chapter 21 • Lithium-Ion Battery Environmental Impacts

Table 21.5

Qualitative Comparison of Battery Recycling Technologies Intermediate Physical

Process

BIT

Process conditions

Combination of physical, Low-temperature, high-temperature, and physical steps chemical steps Co, Li, Al, Cu Cathode, anode, electrolyte, metals Probably Yes

Materials recovered Single chemistry required? Inputs

501

Acid and oxidant

Soda ash

Direct Physical

Pyrometallurgical

Low-temperature, physical steps

High-temperature and chemical steps

Cathode, anode, electrolyte, metals Yes

Co, Ni, Cu No

Make-up CO2, Li2CO3

Limestone, slag, HCl, oxidant

and energy-intensive. In addition, large-scale operation is required. An advantage of physical recycling processes is that multiple high-value battery components can be recovered. This is particularly important as battery manufacturers move away from cobalt-containing cathodes. Current smelters recover much of the value from the battery via the value of the contained elements. However, as can be seen in Table 21.6, which compares the summed prices of the constituent elements to the price of cathode material for several chemistries, the recovery of elements rather than cathode material becomes very unattractive when cobalt content declines. Given the changing battery chemistries on the market and the immaturity of these technologies, further developments in automotive battery recycling technologies are likely. The quantitative analysis of energy use and emissions of recycling technologies, based on calculations outlined by Dunn et al. [8], first examines the energy required to recover LiMn2O4 through these technologies (except smelting, because no Co or Ni is present) and, if necessary, subsequent steps to produce the final active material. Figure 21.13 compares this energy requirement with energy consumed in the production of virgin LiMn2O4 with Li2CO3 from Nevada and Chile. In Figure 21.13, the framed boxes contain the energy associated with the recycling processes. The results indicate that recovery of LiMn2O4 from any of the recycling processes could be less energy-intensive than producing virgin cathode material. In the case of the BIT process, the most energy-intensive Table 21.6

Cathode Material Prices

Cathode

Price of Constituents ($/lb)

Price of Cathode ($/lb)

LiCoO2 LiNi0.3Co0.3Mn0.3O2 LiMnO2 LiFePO4

9.90 6.10 1.35 0.75 [22]

12.00 [19,20] 8.80 [20] 4.50 [21] 9.10 [21]

502

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

FIGURE 21.13 Estimated energy consumption for LiMn2O4 production via automotive battery recycling. Components in framed boxes are produced (Li, Li2CO3 and LiMn2O4) or consumed (H2O2, citric acid and soda ash) in the recycling processes. Other components are consumed during upgrading of recovered lithium compounds to cathode material. (For color version of this figure, the reader is referred to the online version of this book.)

of the processes we analyzed, the consumption of citric acid and hydrogen peroxide is a significant energy burden. Both this process and the intermediate physical process require postprocessing of Li2CO3 with Mn2O3 to produce LiMn2O4, which contributes between 73% and 84% of the total energy consumption of these processes. The cathode material output from the direct physical recycling process requires only minor relithiation. In a closed-loop recycling scenario, combinations of LiMn2O4, aluminum and copper can be recovered from the three recycling processes and incorporated into new Li-ion batteries, achieving the reductions in total energy consumption depicted in Figure 21.14(a). GHG emissions reductions from recycling are displayed in Figure 21.14(b). (Energy consumption values for recycled aluminum include the energies consumed during aluminum scrap melting and casting, sheet rolling and production, and stamping, all from GREET2_2012 [14]. Similarly, to account for fabrication of the recycled copper to a form that could be incorporated back into batteries, we added 50% of the virgin copper production energy [14] to the energy intensity of recovering copper in each of the recycling processes.) The direct physical recycling process, which has only low-temperature steps and recovers the active material directly, offers the most benefit among the three processes. In a closed-loop recycling scenario with this process, nearly half of the total CTG energy consumption of a battery made from virgin materials is conserved when cathode material, aluminum and copper are recycled. Clearly, recovering recyclable metals from

Chapter 21 • Lithium-Ion Battery Environmental Impacts

503

FIGURE 21.14 Total (a) estimated energy consumption (MJ/kg battery) and (b) GHG emissions of BEV batteries made from virgin materials (solid black line); with recycled cathode materials; with recycled aluminum; with recycled copper; and with recycled cathode material, copper, and aluminum. (For color version, refer to the plate section.)

504

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

batteries is important. Recall that these metals are the key contributor to battery CTG SOx emissions; in a closed-loop recycling scenario, these emissions would be significantly reduced. The metal conservation benefit of recovering lithium from the recycling processes, despite the lower energy savings from active material as compared to metals recovery, is also a key point. Additionally, if the cathode were LiCoO2 rather than LiMn2O4, we expect that the benefits of cathode material recycling would increase, because virgin LiCoO2 is much more energy-intensive to produce than virgin LiMn2O4 (see Table 21.3). Although other battery components (e.g. carbon, electrolyte) are recoverable from the three recycling processes, we did not quantify the smaller potential energy, environmental, or economic benefits that might ensue. The results we present here serve as first approximations to the possible benefits of automotive Li-ion battery recycling and likely have significant associated error. These technologies, although patented, are relatively immature, and exact energy consumption and emissions data are unavailable. Additionally, several technical challenges face these processes. For example, it is not clear by what means the electrolyte (including LiPF6) would be recovered by the BIT and intermediate processes, although the direct physical process has demonstrated its recovery at the bench scale. The PVDF binder could prove very challenging to recover. Both of these chemicals could degrade to form halogenated compounds that could pose environmental risks, especially on a local level, if released.

5. Factors that Affect Recycling Numerous factors will come into play to determine whether Li-ion battery recycling will actually be implemented. These can be related to performance and economics of the recycling processes themselves as well as factors concerning health, safety, or the environment. In addition, laws, regulations and government incentives can either enable successful recycling operations or hinder their establishment. This section outlines several prerequisites that must all be met for recycling to succeed. First of all, it must be possible to collect the used batteries at a reasonable cost. In the case of automotive propulsion batteries, there are several likely scenarios. If the batteries are to be replaced at a dealership, they can be accumulated there in large enough quantities for economical collection. For vehicles that have reached their end-of-life, the batteries can be removed and collected before the hulks are shredded. Both of these routes are currently used for lead-acid starting/lighting/ignition (SLI) batteries. In fact, one major US producer takes spent SLI batteries for recycling as a back-haul on its new battery deliveries, but regulations in Europe may not permit this efficient system. Regulations in Europe do require collection of batteries, but high rates have not yet been achieved. Second use of automotive propulsion batteries for utility power storage would likely reduce their environmental impacts and could result in large numbers of batteries at the storage location, and they could be collected from there for eventual recycling. On the other hand, second use at remote locations would make collection difficult. Wherever the batteries are collected and stored, one must ensure that there are no environmental,

Chapter 21 • Lithium-Ion Battery Environmental Impacts

505

health, or safety (such as fire) hazards associated with the storage facility. Safe storage could add to the costs of battery recycling. It may be a challenge to design regulations that will protect the public and the environment without hindering efficient recycling. Collection and processing of damaged batteries is a potential problem that is beyond the scope of this chapter. The batteries that are collected must be determined to be suitable for processing. Today’s battery recyclers must deal with a very diverse feedstock that includes numerous battery types and might even include harmful or dangerous components. Recycling consumer electronic batteries could keep the companies operating until large quantities of automotive propulsion batteries are available for recycling; the automotive batteries may then pose less of a challenge because they will be larger and will probably come in a much smaller number of types or chemistries. Standardization and design-for-recycling could make the job even easier. Recycler requirements depend on the process. Of course, there must be sufficient material to run the process at an economical scale. For pyrometallurgical processing, that means thousands of tons per day, but for direct recycling, only 10 tons per day might suffice. In addition, the material must be sufficiently pure to serve as good input for the process. Pyrometallurgical processing can handle a mixed material stream more easily than can hydrometallurgical or direct recycling processes. If a clean stream is required, labeling to identify battery chemistry, such as that currently being proposed by the Society of Automotive Engineers in the US, would enable sorting into suitable streams for processing. Batteries could be sorted either before or after transport for recycling, either manually or mechanically. Standardization of pack or module shape would enable design of sorting machines for more economical sorting and possibly disassembly. This would be especially beneficial for direct recycling and hydrometallurgical processing. All of the recycling processes require the battery pack to be disassembled, so designing for disassembly would benefit recycling economics. Design for disassembly and recycling can mean minimization of parts, use of reversible fastenings (e.g. nuts and bolts rather than welds), avoiding potting materials that hold parts in place, and minimizing the use of laminated or composite materials that are difficult to recycle. This could mean higher manufacturing costs to enable recycling 10 years later. Note that product performance must remain the top priority, and cannot be jeopardized by recycling considerations. Finally, the key consideration is whether a valuable product can be produced. Whatever the product, it must be of sufficient quality to compete with other raw materials. For pyrometallurgical processes, the product competes directly with metals from ore and is refined to equivalent purity. Products from hydrometallurgical processes can also be purified to be directly competitive with virgin materials. In both of these cases, favorable economics depend on the cobalt content of the battery; current recycling processes are driven by the revenues from cobalt recovery. As cobalt use declines, other incentives will be required to make the business of recycling Li-ion batteries profitable.

506

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

Direct recycling economics do not depend on cobalt content, but do depend on the purity and other properties of recovered cathode material. Further, it is not entirely clear whether cathode material recovered 10 or more years after being put into a battery will still be usable. It might be obsolete because battery chemistry has continued to evolve. In addition, manufacturers in other industries, such as tire manufacture, have often been reluctant to use recycled material because of fears of liability. However, prompt scrap generated by current manufacturers could be recovered confidently by direct recycling today, on a small scale. If a valuable product cannot be produced, recycling must be funded by a regulatory mechanism like a surcharge on product purchases or other fees.

6. Conclusions Preliminary analysis of production of Li-ion batteries with LiMn2O4 cathodes did not reveal any significant environmental impacts or requirements for energy-intensive processing. Detailed analysis of other cathode types is required, but is not expected to reveal significant emission or energy concerns. Battery production impacts are small compared to the use-phase impacts. Local impacts need to be examined to determine if any community experiences a disproportionate burden (e.g. from metals mining). Battery recycling has been shown to reduce energy costs and environmental impacts of supplying Li-ion batteries for vehicles with electric drive. Recycling can also ease supply constraints and reduce the need for imports. Several recycling processes are available or under development. The impact reduction from these processes depends on how far down the process chain recovered materials are located. Since battery chemistry has not yet been decided (and no single winner may emerge), there is an uncertainty that acts as an impediment to development of the recycling industry. In addition, sufficient quantities of feed for large-scale (pyrometallurgical) processing will take many years to build up because the vehicles must first penetrate the market and then go through their useful lives. Smaller-scale recycling processes could operate now on prompt scrap. Economic and regulatory measures imposed by governments could be used to provide incentives for recycling, and care is required to ensure that they do not hinder it.

Acknowledgments This research was supported by the Vehicle Technologies Program in the US Department of Energy, Office of Energy Efficiency and Renewable Energy, under contract DE-AC02-06CH11357. We would like to thank Connie Bezanson and David Howell of the Office of Vehicle Technologies for their support. In addition, we thank several Argonne colleagues for helpful discussions: John Molburg, Kevin Gallagher, Andy Burnham, John Sullivan, and Dan Santini. Finally, we thank Matt Barnes of Pennsylvania State University for his assistance with data collection and analysis.

Chapter 21 • Lithium-Ion Battery Environmental Impacts

507

Nomenclature BatPaC BEV BIT CTG DMC EC GHG GREET HEV NMP PHEV PVDF SLI USGS

Battery Performance and Cost battery electric vehicle Beijing Institute of Technology cradle-to-gate dimethyl carbonate ethylene carbonate greenhouse gas Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation hybrid electric vehicle N-methyl-2-pyrrolidone plug-in hybrid electric vehicle polyvinylidene fluoride starting/lighting/ignition US Geological Survey

References [1] L. Gaines, MRS Bull 37 (2012) 333. [2] L. Gaines, P. Nelson, TMS 2010 Annual Meeting and Exhibition, Seattle, USA, 2010. [3] J. Mervis, Science 337 (2012) 668. [4] European Battery Directive, 26.9.2006 EN Official Journal of the European Union L 266/1(2006). Available from: http://eur-lex.europa.eu/LexUriServ/site/en/oj/2006/l_266/l_ 26620060926en00010014.pdf, (accessed 31.05.13). [5] ISO International Standard, ISO/FDIS 14040, Assessment—Principles and Framework, 1997.

Environmental

Management—Life

Cycle

[6] ISO International Standard, ISO 14041, Environmental Management—Life Cycle Assessment—Goal and Scope Definition and Inventory Analysis, 1998. [7] ISO International Standard, ISO 14042, Environmental Management—Life Cycle Assessment—Life Cycle Impact Assessment, 2000. [8] J.B. Dunn, L. Gaines, M. Barnes, J. Sullivan, M.Q. Wang, Material and Energy Flows in Materials Production, Assembly, and End-of-Life Stages of the Life Cycle of Lithium-Ion Batteries, Argonne National Laboratory, 2012. Report no. ANL/ESD/12–3. [9] Argonne National Laboratory, BatPaC Model, 2011. Available from: http://www.cse.anl.gov/ BatPaC/about.html, (accessed 31.05.13). [10] P.A. Nelson, K.G. Gallagher, I. Bloom, D.W. Dees, Modeling the Performance and Cost of LithiumIon Batteries for Electric-Drive Vehicles, Argonne National Laboratory, 2011. Report no. ANL-11/32. [11] USGS, Mineral Commodity Summaries 2012: Lithium. Available from: http://minerals.usgs.gov/ minerals/pubs/mcs/2012/mcs2012.pdf, (accessed 13.11.12). [12] N. Espinosa, R. Garcı´a-Valverde, F.C. Krebs, Energy Environ. Sci. 4 (2011) 1547. [13] J.B. Dunn, L. Gaines, J. Sullivan, M.Q. Wang, Environ. Sci. Tech. 46 (2012) 12704. [14] Argonne National Laboratory, GREET2_2012 (Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model), 2012. Available from: http://greet.es.anl.gov/, (accessed 31.05.13).

508

LITHIUM-ION BATTERIES: ADVANCES AND APPLICATIONS

[15] G. Majeau-Bettez, T.R. Hawkins, A.H. Strømman, Environ. Sci. Tech. 45 (2011) 4548. [16] L. Li, J. Ge, F. Wu, R. Chen, S. Chen, B. Wu, J. Hazard Mater. 176 (2010) 288. [17] M. Caffarey, Umicore, unpublished data (2010). [18] T. Coy, Kinsbursky Bros., Inc., unpublished data (2009). [19] http://www.asianmetal.com/news/viewNews.am?newsld¼782720. [20] Metal-Pages, 2011, Chinese Prices of Cathode Material for Lithium-Ion Batteries Rise (17.08.11). [21] D. Santini, K. Gallagher, P. Nelson, 25th World Battery, Hybrid and Fuel Cell Electric Vehicle Symposium & Exhibition, Shenzhen, China, 2010. [22] J. Montenegro, T. Ellis, J. Hohn, SAE World Congress 2012: End of Life Management: Reuse/ Recycle/Remanufacture (SDP119), Detroit, USA, 2012.