Decarbonization Potentials for Automotive Supply Chains: Emission-Intensity Pathways of Carbon-Intensive Hotspots of Battery Electric Vehicles

Abstract

To keep global warming below 1.5 °C, the road transport sector must decrease its emissions by substituting internal combustion engine vehicles (ICEV) with battery electric vehicles (BEV). As BEVs can be operated with renewable electricity, the CO_2−eq emissions of the supply chain are relevant for future mitigation. The aim of this paper is to derive emission-intensity pathways and to determine the decarbonization impact regarding the lifecycle emissions of BEVs. Therefore, an analysis for steel, aluminum, battery cells, plastic, and glass, and an evaluation of the literature containing present emission intensities (e.g., for steel 1.7 tCO₂/t to 2.8 tCO₂/t) and reduction potentials, were performed. Based on low-carbon electricity, circular materials, and recycling, as well as technological improvements, emission intensities can be decreased by 69% to 91% by 2050. As a result, the carbon footprint of the reviewed vehicles can be reduced by 47% for supply chain emissions, whereas 25% to 37% of the total lifecycle emissions remain. Considering the scenario studied, BEVs cannot be decarbonized aligned to the 1.5 °C pathway using only avoidance and reduction measures until 2050. Consequently, the application of carbon removals is necessary. However, the applied trajectory and extrapolation relies on material availability and does not consider abatement costs.

1. Introduction

The advance of global warming increases the risk of cascading tipping points temporarily overshooting carbon budgets by more than half [1]. To soften the rise in temperature and to stay within the planetary boundaries, all sectors must decrease their carbon emissions. Thereby, around 34% of the net global greenhouse gas emissions are allocated to the energy sector, followed by 24% to the industry sector, 22% to agriculture, forestry, and other land use (AFOLU), 15% to transportation and 6% to buildings [2]. One main contributor in the mobility sector is the automotive industry as a sub-element of road transport. Thus, the lifecycle emissions of vehicles couple different sectors regarding GHG emissions accounting: based on the upstream supply chain activities required for manufacturing, e.g., steel and aluminum body components, they are connected to the industry sector [3], while the consumption of power for the product use in the downstream value chain is linked to the environmental impacts to the energy sector [4,5]. With the growth of the automotive industry, the decarbonization of every car is required to stay in line with the remaining global carbon budget [6]. Currently, the transportation sector relies mainly on fuels to power combustion engines [7], which emit around 80% of their lifetime emissions during product use as tailpipe emissions [8]. Therefore, electrification is a key technology to reduce total emissions and improve local air quality [9,10]. According to Blas et al. [7], battery electric vehicles (BEV) require three times less the total energy to operate a car than fuel-using drivetrains. Nevertheless, compared to conventional vehicles, the manufacturing of electric vehicles (EV) releases more GHGs, especially by the production of traction batteries [11], and the energy consumption for use depends on technological and behavioral parameters.

This paper compared environmental impacts for different vehicle drivetrains, lifecycle assessments and forecasts on decarbonization pathways for the automotive industry (see background). Many studies focus on the benefits of BEVs in the use phase, reducing emissions with the help of renewable charging and the greening of the energy sector. Often, the remaining emissions, mostly from the manufacturing of car components in the downstream supply chain, are not included in forecasts. As more than one fourth of all lifecycle emissions remain, the analysis of specific value-chain emissions and their reduction potentials must be conducted. To complement the currently used phase-driven mitigation pathway of vehicles, this paper aims:

(1)

to derive emission-intensity pathways for specific materials and components and

(2)

to determine the decarbonization impact of the combined levers for extrapolation.

Therefore, the study is structured as follows: After introducing the field of GHG emissions in the automotive industry (Section 1), the background information deals with insights from the literature and the identification of emissions hotpots for passenger cars (Section 2). Next, the methodological procedure is presented (Section 3). Then, based on specific emission intensities and mitigation potentials, the emission-intensity pathways for the identified emission hotspots are derived and the extrapolation of the upcoming BEV carbon footprint is conducted (Section 4). Lastly, the insights and limitations of this work are discussed (Section 5) and concluding remarks are made (Section 6).

2. Background

Recent forecasts predict that the number of EVs using lithium-ion batteries will be over 220 million by 2030 [12]. Consequently, many studies in the literature deal with the environmental impacts of BEVs, especially their carbon footprint, and comparisons of different vehicle drivetrains.

For example, Ahmadi [6] compared EVs, hybrid electric vehicles (HEV), plug-in hybrid vehicles (PHEV) and internal combustion engine vehicles (ICEV) and determined that BEVs can decrease GHG emissions by half compared to vehicles with internal combustion engines (ICE) due to the possibility of using renewable energy for charging in the use-phase of the products’ lifecycle. In the elaboration from Pipitone et al. [13] a lifecycle assessment (LCA) revealed that the global warming potential of a BEV is around 40% lower than that of an ICE. Nevertheless, the acidifying emissions and particulate matter were higher and the conducted scenario analysis with varying lifetime mileages and different emission intensities of the electricity grid mixes in Europe showed situations where ICEs are favorable [13]. A study was conducted by Hill et al. [14] where over 60 generic combinations of powertrains and body types were analyzed using LCAs with varying electricity and fuel production chains, decarbonization scenarios and sensitivities towards key assumptions and uncertainties for 2020 and 2050. As a result, the work shows that BEVs have lower environmental impacts and smaller deviations due to changing regional and operational conditions today and in the future [14].

In a publication by Wang et al. [15] LCAs were conducted to estimate the CO₂ emissions for BEVs by 2015, 2020 and 2030 in China, integrating specific policies and assumptions that plan to reduce GHGs for electricity generation and manufacturing industries. The calculations revealed a decrease in BEV emissions by 40% compared to those of ICEVs in 2020, which is an improvement to 2015 emissions based on the power mix. For the future scenario, the study determined additional reduction potentials in battery manufacturing and metal production leading to 51 gCO₂/pkm in 2030 [15]. In a study by Fugger et al. [16] a forecasting method was presented using projections for the product portfolio of a fictive automotive manufacturer and decarbonization scenarios to derive a reduction trajectory until 2050. Outlining increasing supply chain emissions and decreasing use-phase emissions due to the transition from ICEV to BEV, the carbon footprint of the company was reduced by 77% by 2050 compared to 2020. This was the result of the decarbonization of the electricity sector, which is responsible for 60% of the reduction, mainly in the use-phase, leaving around 20% of total emissions within the supply chain [16]. The World Economic Forum [17] published an outlook on the decarbonization pathway of passenger cars, quantifying the average CO_2−eq emissions per passenger kilometres in 2020 with 146 gCO_2−eq/pkm. Switching to BEVs with low-carbon electricity for charging, the value could be reduced to 44 gCO_2−eq/pkm in the future, leaving the GHG emissions from material and component manufacturing which should be mitigated by circular materials and undefined innovations [17].

For manufacturing, raw materials are required, and energy is used to process them into products. This leads to emissions in the energy and industry sector. Consequently, Figure 1 shows the distribution of emissions from six automotive OEMs in 2021 [18]. Out of the 17 categories for inventory reporting (scope 1, 2, 3 (cat. 1–15)) [19] it is noticeable that for all the referred companies, two classes are responsible for over 90% of all CO_2−eq emissions: scope 3 cat. 1 “purchased goods and services”, representing upstream value chain activities with suppliers; and scope 3 cat. 11 “use of sold products”, which accounts for emissions during product usage.

In addition to the company perspective, several publicly available LCA (or comparable datasets) have been used to identify the hotspots on the product-level. Illustrated in Figure 2 [20,21,22,23,24,25], the data show the lifecycle emissions of passenger cars divided into three categories, whereas each model is described with two different bars. The first bar describes the total emissions with the EU energy mix and the second one the remaining GHG emissions when renewable charging (e.g., based on hydropower) is applied. Using the decarbonization measure for the electricity generation in product usage, average emissions can be reduced from −30% to −53%. As these comparisons confirm the lifecycle hotspots on the company and product levels identified in the present literature, it is also derivable why this elaboration focuses on the remaining emissions in the manufacturing stage that cannot be avoided by one single measure.

A passenger car requires many different materials, from metals like steel, aluminum, and copper [26] to plastics, rubber, glass, and rare earth elements [27]. Additionally, raw materials such as lithium, nickel and cobalt are needed for the battery, and gold and silicon are required for electronic components [3]. To determine which of the relevant materials and components are the major emitters, an analysis of the material composition and their GHG distribution within the upstream value chain was conducted. Figure 3 presents exemplary sources of emissions and their contributions to the absolute CO_2−eq emissions [14,20,21,22,28]. Depending on the size and model of the car, steel and light alloys comprise the largest share in a vehicle [21]. Plastics, non-ferrous metals and glass comprise the second-largest share [20]. However, not all materials with a high share in weight are responsible for a significant percentage of global warming potential. In fact, the battery cells are the major emission hotspot for supply chain parts (28% to 31%), corresponding to data from VW [22]. Looking at BEVs without their drivetrain impact, steel and aluminum are the second and third emissions contributors, whereby the relation can vary as premium models use more aluminum to increase their performance. Lastly, textiles and plastics are responsible for around 28% of the remaining emissions [14].

3. Methodology

Shown in the background section, the present research focuses on drivetrain comparisons with emission reduction potentials based on the electrification and renewable charging. Nevertheless, the comparison of corporate and product-specific emissions illustrates that the use-phase and the supply chain are major sources of emissions along the lifecycle for automotive OEM. Moreover, evaluating LCA data outlines hotspot materials like steel, aluminum, battery cells, plastics and glass within value chain emissions. Consequently, these components are focussed in this work to analyze their future mitigation trajectories.

To fulfil the research objectives, the work follows the procedure shown in Figure 4. Therefore, for each selected hotspot in the supply chain an industry analysis is conducted which includes the gathering of CO_2−eq emission-intensity data, the emission distribution for the manufacturing process and GHG mitigation potentials. Based on the information, individual emission-intensity pathways are derived. Then, the trajectories are used to determine an extrapolation of automotive lifecycle emissions to calculate the decarbonization impact of all measures along for BEVs. Lastly, the resulting and remaining amount of emissions is set in relation to scientific requirements for 1.5 °C aligned decarbonization scenarios and automotive mitigation pathways.

For each material, the first step of the analysis is the gathering of emission-intensity data. Therefore, present literature is evaluated regarding their determinations and calculations of CO_2−eq data. For the main emitters, steel, aluminum, battery cells and plastics, at least 30 intensity values are collected to show the range of varying numbers based on spatial and technological differences. Moreover, a global, European, American or Chinese average value is added to set a baseline. When no specific data are applicable, the average baseline is calculated using all gathered samples. However, an exception for this procedure is made for glass, as it is a minor contributor.

In the second part of the industry analysis, the GHG emission data is used to assess the distribution of CO_2−eq. In order to find mitigation potentials and reduction levers in the following step, the emission distribution is outlined to highlight manufacturing or raw material hotspots. Thirdly, decarbonization measures are investigated. Evaluating different publications for each component, activities to improve the environmental impacts are identified. Thereby, avoidance and reduction potentials were prioritized against removal measures although they could be required to perform deep decarbonizations.

After the industry analysis is conducted and data are collected, the derivation of emission-intensity pathways is next. For the calculation of each trajectory, the identified mitigation levers are combined with the CO_2−eq intensity sets under specific conditions and assumptions. Initially, the baseline (average) GHG emission value is complemented by the upper and lower boundary of the published numbers to demonstrate the deviations that can occur by regional differences or technological properties. Moreover, the implementation of measures is based on the sector trajectories of the IEA NZE 2050 scenario, where the energy sector is decarbonized by 2040 [29]. This determines the grid mix developments and the applicability of renewable power. Another specification is the availability of scrap and raw materials which is required to forecast the use of all measures without limited constraints. In reality, shortages can occur and have a primary impact on costs and feasibility [7].

Lastly, based on the calculated emission-intensity pathways an extrapolation regarding the decarbonization impact of all levers is performed in the last step. Thereby, each reduction potential of the material-specific trajectories is weighted based on the share of emissions and weight for a BEV.

4. Results

In accordance with the presented approach, the identified hotspots for GHG emissions in the supply chain are analyzed. Therefore, the manufacturing stages and decarbonization measures for the relevant materials and components are evaluated. The gathered data is used to derive emission-intensity pathways and to extrapolate the decarbonization impact.

4.1. Emission-Intensity Pathways

For the identified supply chain emission hotspots including steel, aluminum, battery cells, plastics, and glass, a multi-level industry analysis was conducted to review current CO₂ intensities, reduction potentials and measures. Afterwards, by linking the present intensities with possible levers for decarbonization, mitigation pathways were derived.

4.1.1. Steel

The manufacturing of steel can be mostly divided into two production routes, where the blast furnace-basic oxygen furnace (BF-BOF) is responsible for around 70% of global steel production and almost all primary materials [30]. The other main route is the electric arc furnace (EAF) which produces around 29% of steel with a high share of secondary materials [31].

For the BF-BOF application, iron ore is reduced with coke to iron before the required carbon level is created and other metals are added for the required material properties [32]. In the EAF process, reduced input material like scrap metal, pig-iron or direct-reduced iron (DRI) are melted with electric currents [33]. BF-BOF manufactured output is produced as flat steel, commonly used in automotive applications based on high-quality properties, while scrap/EAF steel is created for the long products mostly used in construction [34]. The energy demand and GHG emissions are determined by structural parameters like scrap availability, the regional power mix and energy supply, the lifetime, and the used technologies [34]. Based on this, the different manufacturing routes should be compared against the same technologies across different countries, as, for example, the average emission intensity of a country can be misleading [35]. Hotspots for the BF-BOF process are tblast furnaces with 61%, and coke-making plants, with 27% of total CO₂ emissions [36].

For EAF, the electricity demand for smelting is responsible for the most GHG and depends on the spatial grid mix [32,37]. Figure 5 [26,30,32,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51] shows the comparison of emission intensities for steel manufacturing routes where the global average is at 1.9 tCO₂/t. For the main contributor BF-BOF, the values found in the literature vary from 1.7 tCO₂/t to 2.8 tCO₂/t (3.5 tCO₂/t), and for EAF the range is 0.12 tCO₂/t to 1.2 tCO₂/t. Alternatives are also included that are relevant for their decarbonization potentials. Therefore, the DRI-NG route has an emission intensity of 1.3 tCO₂/t to 1.4 tCO₂/t, the DRI-EAF 0.76 tCO₂/t to 1.5 tCO₂/t and smelt reduction (SR) has an emission intensity of 1.35 tCO₂/t.

The increase of scrap for the production of steel is one lever to reduce CO₂ because for each ton of steel produced, around 1.4 t of iron ore, 0.8 t of coke and 0.3 t of limestone can be saved, leading to a decrease of 1.67 tCO₂ [98,99,100,101]. However, for BF-BOF only a share of the 30% of scrap is possible (present ~20% [26]), while for EAF the high scrap share can influence material properties [32].

Second, the increase in energy efficiency for conventional manufacturing can lead to a decline in emissions of between 25% to 40% [42]. Moreover, biocarbon and biomass can be used to decarbonize BF-BOF processes. This results in a reduction potential of 43% [37]. For EAF, operation using renewable energy is one decarbonization lever, as emissions can be decreased to 0.1 tCO₂/t [34].

Another option is the implementation of new low-emission reduction routes like (H)DRI-EAF/DRI-NG [102]. The third production pathway can be based on natural gas [34] or renewable hydrogen as reductant for the iron [36,38]. The application of H₂ can lead to a decrease of 47.5% to 95% depending on the hydrogen generation used [37]. Additionally, the DRI-process can be combined with scrap for the melting bath in the EAF process, which leads to an increased scrap share of 50% to 60% [26]. However, the energy demand of this process route is significantly higher and consequently a low-carbon energy supply is required [36]. Moreover, Electrowinning (EW), as one electrolysis route, could be an alternative as the iron ore is suspended in a solution at 110 °C to generate iron. Thereby, the chemical process does not emit carbon dioxide and the required power demand can be supplied by renewables [26]. A fifth route is smelting reduction, where iron ore is transferred in a reactor to be liquified and then reprocessed to steel in a BOF leading to energy savings of 20% [26].

Lastly, carbon capture and storage (CCS) can be applied for all technologies to abate the remaining emissions [36]. For BF-BOF, an analysis was conducted that quantifies the impact of integrated CCS to between 48% to 76% [37,103].

4.1.2. Aluminum

The second material-based emission hotspot is the use of aluminum for body components, battery casings and the rims of BEVs. Moreover, the element is applied to partly substitute mild- or high-strength steel for lightweight applications to save weight [104]. As a consequence, recent studies show that the demand will triple until 2050 [53]. To produce aluminum there are two manufacturing routes [105]: the primary path, which uses bauxite, or the secondary path using scrap [61]. To fabricate the end-product, aluminum oxide is made from bauxite through the Bayer process [106] and then the Hall–Heroult process is used for electrolytic reduction to create aluminum [52].

The primary manufacturing of aluminum is one of the most energy-intensive industrial processes [52]. Mining, refining, anode production and smelting account for around 90% of all CO₂ emissions [56] whereas 65–72% of these GHGs are related to electricity supply [54,58,106]. The power is mainly required for the Hall–Heroult process where around 20% of total CO₂ emissions (and perfluorocarbons (PFCs)) are a result of the anode effect [52,54,61].

Figure 5 [43,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66] compares the emission intensities of produced primary aluminum with secondary materials and average industry values that are not specified in detail. An intensity of 14.6 tCO₂/t is declared as a global mean value by the International Aluminum Institute. However, primary materials vary from 5.4 tCO₂/t to 21.7 tCO₂/t (41.1 tCO₂/t) based on country-based energy mixes and fuel supply. For secondary materials, the analysis shows values from 0.32 tCO₂/t to 2.3 tCO₂/t.

The first decarbonization potential for aluminum is foreseeable, as the review of emission intensity demonstrates that secondary materials have a lower carbon footprint than aluminum produced on the primary route with a high-carbon energy supply [107]. Today, more than 90% of all aluminum parts in the building and automotive industry are recycled but the rates can vary due to products, alloys, and regions [53,58]. Around 95% of the energy demand and 92% of CO₂ emissions can be saved if scrap is used [43,98]. However, scrap is limited [105].

Based on high energy demand, substitution with renewable power like hydropower for the smelting process can reduce emissions [38,53,105]. A third lever for GHG reduction is the replacement of the carbon anode with an inert anode which avoids direct emissions from the Hall-Heroult process due to conversion of O₂ instead of CO₂ [57,105,108]. Nevertheless, the inert anode smelting technology requires 50% more energy compared to the conventional procedure in the theoretical optimum (~6.2 kWh/kg vs. ~9.2 kWh/kg) which makes the alternative only beneficial when low-carbon energy is used [57,58,109]. In addition to an improved anode, a wetted cathode can also improve the environmental footprint of the Hall–Heroult process, leading to reduced emissions of around 10% to 23% [53,57,61,105].

Using the boehmite process in the first production step to create aluminum oxide from bauxite, energy demand can be reduced by 19%, correlating with decreased CO₂ emissions of 22% [52]. Regarding process adaptions, chemical carbothermic reduction is an alternative to the Hall–Heroult process even if the industrial readiness is pending. Thereby, carbon is used as a direct reduction agent instead of the present carbon anodes, leading to an improvement of at least 23% with coal-powered energy [52,110].

4.1.3. Battery

For electric vehicles, lithium-ion batteries (LIB) are the current market standard based on their high energy density, low self-discharge and long lifecycle [67,111,112]. LIBs can be differentiated by different compositions of active cathode materials with different properties [74,113]: nickel manganese cobalt oxide (NMC), nickel cobalt aluminum oxide (NCA), lithium iron phosphate (LFP), lithium manganese oxide (LMO) and lithium cobalt oxide (LCO) [69,74,114]. However, they are the origin of more than one third of BEV lifecycle emissions [80,115].

After the extraction of raw materials, which are used for the cathode and are responsible for up to 80% of GHG emissions [81], battery production contains processes like homogenization, coating, drying, rolling, die-cutting, packaging, liquid injection and aging [67]. Within the assembly process, coating and drying are the most energy-intensive steps, with 35% to 40% of the total energy consumption [11,67]. From the remaining emissions, the aluminum casing and the anode are the subsequent CO₂ hotspots [67].

Figure 5 [3,11,12,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85] shows the emission intensities of different batteries by cell chemistry found in the literature. Since 1999, there have been more than 100 studies dealing with the topic of environmental impacts of LIBs published [73]. Nevertheless, existing articles present results that range from 36.9 kgCO₂/kWh to 487 kgCO₂/kWh depending on cell chemistry, electricity mix, battery lifecycle, inventory data for upstream materials and processes including transparency and benefits from remanufacturing [67,71,73]. In this work, the average value of all samples is 114.4 kgCO₂/kWh.

Reduction potentials for lithium-ion batteries concentrate on renewable energy supply, increased production efficiency, the use of recycled materials and adapted cell chemistries [67]. Comparing cell chemistries, recent studies reveal that LIBs are advantageous against Ni-H, lead-acid, or Ni-Cd batteries [116,117,118]. Within LIBs, LFP batteries have a reduced carbon footprint of around 42% to 45% compared by weight but based on the lower energy density, the benefit can be mitigated [75,119,120]. For future applications, solid-state batteries (SSB) are outlined as an innovative approach to reduce GHG emissions [121]. Studies show that SSBs using an oxide electrolyte and an NMC cathode can reduce the carbon footprint by 24% but will shift the emission hotspot to lithium as they require 35% more of this metal [74,75,122,123].

Switching the energy supply to renewable electricity could decrease the total carbon emissions by 49% to 76% as the cathode production and assembly emissions are reduced [12,76]. An additional lever for decarbonization is the increased production efficiency for battery assembly as the production sites must stay in energy-intensive conditions (vacuum, constant temperature). When the number of finished pieces per minute (ppm) can be increased, the fixed energy demand can be distributed to more modules. Consequently, an increase from 30 ppm to 40 ppm and to 50 ppm can improve the CO₂ balance of the assembly stage by 25% to 40% [67].

The last measure to avoid emissions is the implementation of secondary materials using recycling processes to recover resources. Therefore, physical, pyrometallurgical and hydrometallurgical methods can be applied, where the reduction potential varies from 4.8% (pyrometallurgical) to 33.5% (hydrometallurgical) to 43.9% (physical). However, these processes require material reproduction, emit secondary pollution and are not applicable at an industry scale currently. [67]

Combining the presented measures will lead to a reduction of >85% with a low-carbon energy mix and secondary materials, which requires further activities in carbon removals for net carbon neutral modules [67].

4.1.4. Plastics

The chemical industry, relying on petroleum and gas [124], has the highest energy demand. One of its products, plastics, can be used in various applications based on their different types and properties [125]. Using unrefined petroleum, flammable gas, or coal, plastics are manufactured using polymerization processes [125].

Regarding emissions intensities, Figure 5 [86,87,88,89,90,91,92,93,94] displays the current values in the literature for virgin (PP, PET, PVC, PUR, HDPE) and recycled plastic, the recycling process, bio-based plastics (corn-based, sugarcane-based) and substitutional biomaterials (flax, hemp, kenaf, jute). With a European average of 2.9 tCO₂/t, the lowest carbon footprint can be found at 1.3 tCO₂/t for HDPE and the highest at 5.7 tCO₂/t for PUR. Nevertheless, biomaterials undercut emissions with numbers below 1 tCO₂/t, but to ensure the required properties for the automotive industry, components must be reinforced, leading to additional emissions [89].

For plastics, the following mitigation strategies are identified: renewable energy, secondary materials, bio-based materials and carbon capture and use (CCU) [91]. Although plastics are declared as a circular material, only 15% of plastic waste is recovered annually for new applications [88]. For recycling, mechanical, and chemical technologies can be applied where either the end-of-life plastics are shredded into flakes or granule forms or broken down into their monomers before they are re-polymerized [86,87,126,127]. However, the recycling processes emit additional CO₂ for processing, waste collection, and sorting, which can decrease the environmental benefits [86]. All in all, a case study calculated a reduction potential of up to 75% for carbon emissions, whereas mechanical recycling emitted the lowest process footprint compared to chemical recycling and energy recovery [86].

The second approach to reduce the lifecycle GHG emissions of plastics is low-carbon renewable energy. An analysis by Zheng et al. [91] built a scenario for US plastics production where the CO₂ emissions were reduced by 50% to 75%. In this case, the energy demand of industrial processes for heating based on fossil-fuel combustion can be substituted with low-carbon electricity and the steam-cracking process can be electrified with the help of green hydrogen [124].

Alternatives to conventional plastics include bio-based plastics and biomaterials. Using renewable raw materials like corn or sugarcane can reduce GHGs by 15% to 25% in specific applications with comparable properties [91]. The combination of all the presented decarbonization potentials including a share of 70% secondary materials can lead to net-zero emission plastics when CCU is used for the remaining non-abatable emissions [128].

4.1.5. Glazing

Among other applications, glass is used for buildings, in vehicles and in photovoltaics [99]. For the production of glass, sand is heated in furnaces and turned into hot glass which then can be molded [129]. In detail, limestone and soda ash are added to the process and a temperature of around 1500 °C must be reached for melting [130]. Consequently, the combustion of fossil fuels like natural gas is responsible for around 50% to 85% of the total emissions [95]. The remaining emissions are process-related due to the decomposition of soda ash and limestone [131].

In Figure 5, [87,95,96,97] the identified emission intensities for different types of glass can be found. Depending on the application case and properties, the values vary from 0.36 to 0.76 tCO₂/t including a European average of 0.46 tCO₂/t, an American average of 0.54 tCO₂/t and a Chinese average of 0.69 tCO₂/t. The difference is a result of the used fuel mix and the share of secondary materials [96].

Based on the research for decarbonization potentials, the analysis shows that the increase of energy efficiency in heating processes, a low-carbon energy supply, and the use of recycled materials are the main levers to reduce the environmental impacts of glass production [38]. The first measure is recycling, where today only 11% of flat glass is recycled globally [99], although glass can theoretically be recycled indefinitely [87]. However, using cullet, which is crushed recycled glass, can decrease the required furnace temperature, avoid process emissions, and reduce landfill, leading to savings of around 5% CO₂ per 10% of cullet mass used in manufacturing [130]. As a result, the European Container Glass Federation (FEVE) quantifies the total impact of 100% cullet supply with around 58% fewer carbon dioxide emissions [131].

Another way to improve the environmental performance of glass manufacturing is the implementation of waste heat recovery for preheating, with reduction potentials of up to 15% [132]. The third identified option is the electrification of furnace technology, e.g., using plasma melting or electric-arc furnaces. With these technologies, the energy demand can be decreased, and low-carbon energy applied, resulting in 20% to 80% reduced CO₂ emissions [129,132].

4.2. Trajectory

After the industry analysis is conducted and for each emission hotspot present emission intensities and decarbonization potentials are identified, CO₂ development pathways can be derived. Based on the average emission values as well as the upper and lower boundaries of the review, the baseline of 2020 was determined. Then, all measures were considered, including certain constraints regarding technology readiness, using assumptions like the grid-mix developments described in the IEA NZE 2050 scenario [29] and no shortages on scrap availability. Figure 6 displays the emission-intensity pathways for steel, aluminum, battery, plastic, and glass. Moreover, required sector trajectories aligned with the aim to limit global warming to 1.5 °C were used to frame the predicted performance [133]. As a result, the illustration shows that steel could reach an intensity of 0.6 tCO₂/t by 2050 compared to 1.9 tCO₂/t in 2020 (−69%).

The reduction is based on a low-carbon electricity supply, increased amounts of scrap, the switch to DRI-NG and the ramp-up of (H)DRI-EAF as well as gains in energy efficiency, and the use of biocarbon and CCS for BF-BOF. For aluminum, the transition from 14.6 tCO₂/t to 2.2 tCO₂/t (−85%) relies also on renewable power, the implementation of inert anodes, more secondary materials and possible integrations of the boehmite process and carbothermic reductions in future. The decarbonization pathway of batteries, where the emission intensity varies widely, uses optimized and innovative cell chemistries, recycled raw materials, improvements in the battery assembly and green energy to decrease emissions from 114.4 kgCO₂/kWh to 10.3 kgCO₂kWh (−91%). Plastics will profit from the grid-mix transformation, increased circular economy and new materials, which can reach the required properties in the future. This results in a total reduction potential of 76%, decreasing the average emission intensity from 2.9 tCO₂/t to 0.7 tCO₂/t. Lastly, the analysis for glass shows options to reach 0.1 tCO₂/t (−80%) with the help of improved furnace technology, waste heat management, and cullet.

4.3. Extrapolation

The background information highlighted the product-use phase and the manufacturing stage (upstream value chain) as major emission hotspots in the automotive industry. Moreover, lifecycle emissions were reviewed in detail based on LCAs from automotive OEMs for specific vehicles. Thereby, all manufacturers showed the potential to decarbonize the CO_2−eq emissions during the use phase with the help of renewable charging. Consequently, the carbon footprint will decrease by 30% to 53%, leaving the manufacturing hotspot for the remaining decarbonization efforts. For this lifecycle stage, an industry analysis was conducted to identify levers for decarbonization and derive emission-intensity pathways. The presented results in Section 4.2 were used to extrapolate the average vehicles emissions in 2050 when all reduction potentials are implemented. Considering that the five supply chain hotspots are responsible for around 60% of all emissions, the results of extrapolating the combined mitigation impact from 2020 to 2050 are shown in Figure 7. Subsequently, the upstream value chain emissions can be reduced by 47% in 2050. This leads to total remaining emissions of 37% to 25% for a vehicle lifecycle.

In order to fulfil deep decarbonization and claim net-zero GHGs for the automotive industry, lifecycle emissions must be decreased to 10% or below of the baseline emissions, as no more carbon, technical, or nature-based removal measures are allowed to compensate non-abatable emissions in the frameworks of international organizations [134]. As a result, at least 15% to 27% of the residual emissions must be abated with additional measures for other materials and components.

5. Discussion

The decarbonization of supply chain emissions in the automotive industry becomes more relevant with the transformation of ICEVs to BEVs. In addition, the development of a low-carbon use phase is available with the growth of a charging infrastructure that prioritizes renewable electricity. This field of activity shifts the focus for climate mitigation to the upstream supply chain where emissions are determined by raw-material extraction and fabrication processes. This study outlines four major levers to reduce the environmental impact using low-carbon energy, recycled and secondary materials, implementing technology innovations in manufacturing routes, and applying carbon removals for the remaining emissions.

However, the demand on key resources and recycled materials can result in market conflicts that may lead to limited availabilities and shortages [7]. Moreover, OEMs only have direct control over their first-tier suppliers, whereas decarbonization objectives must be aligned and communicated over the whole value chain. This can cause conflicts of interest as the requirements for low-carbon components increase while new investments to reduce emissions are not always financially covered to ensure stable prices [135]. Furthermore, especially in the heavy industry, production plants and machines have technical lifetimes of up to 50 years (e.g., blast furnaces in Germany) which can slow down the green transition as stranded assets are already planned and lock-in effects are present [34]. Based on the obstacles regarding the main decarbonization levers, a slower energy transition compared to the normative IEA NZE 2050 scenario can reduce the availability and implementation of renewable power for mitigation. Furthermore, when not enough secondary materials are available in global markets, the recycling potential and CO_2−eq savings are delayed and the costs can become prohibitively high. Lastly, the long lifetimes of industry facilities lower the conversion to innovative and more efficient technologies.

From an overall perspective, the outcome of the extrapolation can be compared to requirements derived from decarbonization scenarios that are aligned to the 1.5 °C global warming target and the set reduction pathways for the automotive industry. Thereby, the CO₂ emissions of cars must be reduced by 94% in 2050, which leads to a remaining target gap of at least 19% [136]. Moreover, comparing the required decarbonization pathways from the Transition Pathway Initiative (TPI) for steel and aluminum with the results of this study, the measures are insufficient to be aligned with the requirements. Nevertheless, this work focuses only on five GHG contributors, and other studies that are considered in the introduction disclose additional improvements for other components that profit e.g., from energy transition, which could decrease the demand of pending reduction activities [16].

Yet, the responsibility to develop low-carbon road transport cannot be done alone by the automotive industry. Climate targets are announced by cities, nations and groups of states covering all emitting sectors. Therefore, these institutions must regulate the environment and create option spaces for improved acting and collaboration. For example, the European Union tightened their goal to reduce GHG by 2030 using the European Trading Scheme, which is based on carbon taxes, the Effort-Sharing Regulation and the Renewable Energy Directive (RED) to affect the energy, industry, transportation, building and AFOLU sector [137]. However, the overarching objectives are left for implementation and present forecasts show insufficient progress [137].

In addition to cross-sector regulations, specific policies are under development, such as the EU legislation for batteries and waste batteries. Consequently, minimum levels for recovered materials and recycled content for use in new batteries are determined to reduce the primary material demand and enhance a circular material that reduces the costs from a long-term perspective [138]. Moreover, the main driver for BEV prices must decrease the costs as predicted from >1000 €/kWh in 2010 to <100 €/kWh in 2050 to roll out electromobility as a valid substitution for global individual mobility [139].

Finally, present activities in the field of manufacturing (Industry 4.0) and digital product services (internet and communication technology) can have causal effects on the environmental performance of upcoming vehicles as energy demand grows faster than the capacities of renewables or efficiency savings [140].

The elaboration is limited by several circumstances. One critical aspect is the data quality of the reviewed and used emission intensities for the materials and components. For the industry analysis and comparison of carbon footprints, published values are used which cannot be directly assigned to manufacturing processes or included properties. Consequently, they are declared as undefined or average. Moreover, it is not ensured that all numbers for CO_2−eq intensities are based on the same assumptions and parameters. Additionally, the impact of reduction measures can include other substitutional effects based on overlapping levers. Another risk is based on the spatial variations regarding manufacturing routes and production conditions, as the baseline values do not consider regional differences and generalize the decarbonization impact on a global level.

Lastly, the work is restricted due to a focus on the five main emission contributors of battery electric vehicles. From the passenger car perspective, all components must be evaluated for reduction potentials. Additionally, the decarbonization of the road transport sector is framed in this approach, as the transition from ICEVs to BEVs is proposed as the main lever and focus although transportation in general can rely on other options. For example, ridesharing or pooling, the use of public transfer by rail or the improved infrastructure for bicycles can lead to a decrease in CO₂.

This leads to future research directions regarding this work: for future elaborations, the work could be extended to all carbon emitters within the supply chain of BEVs. Furthermore, the data quality of all numbers can be increased by using only comparable values with the same scopes. Moreover, company-specific values for improvements could be gathered and used in addition to theoretical and case-study-based information.

Beyond the input parameters to derive emission-intensity pathways and determine the decarbonization impact, the outcome could be analyzed in complex system dynamic simulations to review the interdependencies of measures, the demand in resources and the connection to cost forecasts.

Finally, the results from for the automotive industry could be transferred to other participants in the road transport (e.g., heavy-duty trucks) sector or other products and industries with overlapping supply chain components.

6. Conclusions

Based on the increased requirements for the transport sector to decrease annual GHG emissions, the portfolio transformation from ICEVs to BEVs is constantly accelerating. Applying the decarbonization levers that are built on low-carbon electricity, the use of circular materials, recycling, bio-based materials, and biomass as well as technological process improvements and innovations for steel, aluminum, battery, plastic and glass, reduction potentials were predicted.

As a result, steel emissions intensity can be decreased by 69% from 1.9 tCO₂/t to 0.6 tCO₂/t by 2050. Second, aluminum can reach a medium CO₂ intensity of 2.2 tCO₂/t from a global average of 14.6 tCO₂/t in 2020. For battery cells, a reduction from 114.4 kgCO₂/kWh to 10.3 kgCO₂/kWh, which equals −91%, is forecast. Lastly, plastics and glass should lower their emissions by about 76% and 80%. Consequently, the mean carbon footprint of the reviewed vehicles from Mercedes-Benz, Volkswagen, Volvo, and BMW can be decreased to 25% to 37% of their total lifecycle emissions. This means that, with the present inventory data, reduction measures and innovative technologies, BEVs cannot be decarbonized aligned to the 1.5 °C pathway using only avoidance and reduction measures until 2050. Consequently, the application of carbon removals is needed to ensure compliant mitigation for the transport sector and to reach a state of deep decarbonization with <90% of baseline emissions compared to 2020 unless new technologies can be applied or the composition of materials changes. However, many of the presented measures can already be implemented today, leading to significant reduction potentials for the automotive industry.

To achieve the objective of climate neutrality on company, national and international levels by 2050 or before, all stakeholders must work together to eliminate obstacles. Subsequently, policy regulations need to address the required emission-intensity pathways for industry goods and lifecycle emissions of vehicles. Moreover, low-carbon roadmaps with distinct milestones must be implemented, carbon prices established, and the energy sector must transition to 100% renewables by 2040 according to normative scenarios. Moreover, minimum levels of recycled content must be introduced and steadily increased through the use of regulations.