1. Introduction
The Korean automobile industry is a critical sector that drives production, value creation, and employment. In 2017, it contributed significantly to the manufacturing sector, accounting for 12.9% of production output and 10.1% of value creation [1]. In 2020, it commanded 12.7% of manufacturing output and 12.1% of exports, while employing 7.1% of the workforce [2]. This shows that the automobile industry is intricately intertwined with the extensive supply chain and holds a pivotal role in the Korea economy. However, the Korean automobile industry faces a pressing challenge to meet the national and global trend of carbon neutrality in 2050 triggered by Paris Agreement. The transportation sector has been reported to contribute 25% of global CO2 emissions and major economies have tightened transportation-related GHG emission regulations, particularly in the automotive sector (e.g., US EPA’s SAFE standards and EU’s 2035 combustion engine ban) [3, 4]. Due to the implementation of these regulations, the leading car manufacturers have been focusing on fuel efficiency in operation phase and transitioning from internal combustion engine to electric vehicle (EV) and securing their low carbon supply chain to comply with GHG and environmental regulations [5].
Life cycle assessment (LCA) serves as a method to quantify carbon emissions across the entire product lifecycle (cradle-to-grave) and encompassing the intricate web of the supply chain. The role of LCA in automobile was highlighted in the WP.29/GRPE Workshop on carbon Life Cycle Assessment (LCA) of vehicles held in May 2022 and the European Commission evaluated the applicability of LCA to automotive GHG emission standards and report to the European Parliament [6]. Relatedly, Japan and US are working to establish an international standard for automobile LCA tool [6]. To quantify the life cycle GHG emissions (carbon footprint), the bottom-up process-based LCA (pLCA) and the top-down input-output-based LCA (IO-LCA) have been applied. The strengths and limitations of the two methods are widely described. The pLCA is detailed but could lead to serious truncation errors for incomplete system boundaries [7, 8]. Meanwhile, the IO-LCA is relatively complete but is aggregated, difficult to be detailed and the aggregation errors could outweigh the truncation errors under certain situations [9, 10]. There are some studies on automobile LCA. Buberger et al. compared the total life cycle GHG emission of commercially available passenger cars with different energy sources [11]. In another case, Pipitone et al. compared the environmental impact of internal combustion engine vehicles (ICEVs), hybrid electric vehicles (HEVs), and battery electric vehicles (BEVs) [12]. However, these approaches haven’t addressed the characteristics of emissions such as Scope 1, 2, and 3 over the entire supply chain; there remains an unsettling void—a glaring absence of systematic analysis about the automobile’s life cycle GHG emissions. The newly adopted Carbon Border Tax (CBT) of the EU and the Inflation Reduction Act (IRA) in the USA and the EU are strengthening supply chain emissions of Scope 3 for the product carbon accounting [13–17] Thus, all companies have to prepare to reduce their supply chain emissions of the product, not just within the manufacturing process.
Given these situations, this study aims to fill the research gap by analyzing the comprehensive GHG emissions from automobile production in Korea. Automobiles, with their intricate comprehensive picture of tens of thousands of components—steel, aluminum, rubber, glass, plastics, paint, and more—present a formidable challenge to the pLCA approach [18–20]. Thus, we navigate this labyrinth with a calculated choice—the IO-LCA, represented by the interdependence between economic interactions and environmental loads. We have calculated GHG emissions based on the energy sources purchased by each industry and weighted averages of carbon emission factors of energy sources.
Our group developed the Korean EEIO table for the year 2017 for analyzing supply chain GHG emissions in Korean economic activity [21, 22]. By using same approach, this paper conducts a thorough analysis of the Korean automotive industry, focusing on GHG emissions, and compares both direct emission (Scope 1), emission from the use of electricity and heat (Scope 2) and the remaining emissions from supply chain (Scope 3). The key research questions of this research are: 1) What are the differences between direct and total GHG emissions, 2) What are the Scope 1, Scope 2, and Scope 3 in the automobile sectors? 3) What are the contributors and drivers of the total GHG emissions? The result will provide crucial insights for the Korean government and automobile companies to establish scientific and practical carbon mitigation strategy.
2. Methods and Data
2.1. Research Framework
Fig. 1 shows a framework for the sequence of steps in the research process undertaken in this study. The first step is the formulation of the EEIO table through the integration of input-output table (IOT), and environmental satellite table developed by energy balance (EB), fuel GHG, and carbon emission factors, along with national GHG inventory data. The second step is EEIO analysis using the EEIO model to estimate the direct and indirect GHG emissions in the automotive industry, delineating emissions by Scope. Additionally, it identifies the contribution of the supply chain to total GHG emissions based on final demand. Step 3 presents the results of the carbon footprint analysis in automobile industry in 2017. Step 4 comprises the conclusion, encompassing key findings, contributions, and the acknowledgment of limitations.
2.2. Data Collection
Our EEIO model is built upon five primary data sources: 1) base sector IOT in 2017 (381 sectors reclassified), provided by the Economic Statistics System under the Bank of Korea [23]; 2) an EB compiled by Korean Energy Economics Institute by 27 energy sources in the same year [24]; 3) Korea’s 2017 GHG inventory compiled by the GHG Inventory and Research Center under Korean Ministry of Environment [25]; 4) the carbon emissions factors for energy sources [26, 27]; and 5) the currency conversion value from Korean Won (KRW) into United States dollar (USD, $). In 2017, a conversion rate of 1129.0435 KRW/$ was used [28].
2.3. Environmentally Extended Input-Output Table Compilation
The structure of the EEIO table is shown in Fig. 2. As shown in Fig. 2, the EEIO table summarizes the amount of energy consumption by industry and GHG emissions by industry according to economic activities, which can analyze the direct and indirect ripple effect of GHG emissions. When the final demand for each industry’s inputs is known, GHG emissions directly or indirectly induced in the production process of each industry to meet this final demand can be estimated, and environmental impacts linked to these economic activities can be analyzed.
Following the previous studies, we compiled the GHG satellite table [21, 22, 29, 30]. First, CO2 emissions for different energy sources consumed by aggregated sectors in the EB are calculated using corresponding emissions factors. For some energy sources, a carbon storage factor fraction is applied. Thus, fuel CO2 emissions in Kiloton (kt) was calculated by Eq. (1):
where NAij is the fuel use (1000 TOE), FCSij is the carbon storage fraction (%), 41.868 is the joule-toe conversion coefficient (TJ/1000 TOE), CFi is the conversion coefficient (net calorific value/total calorific value), EFi is the emissions factors (tC/TJ), 44/12 is the conversion rate from carbon to carbon dioxide (kg CO2/kg C), i is fuel type, and j is sector. The sectors in the EB often consume multiple types of fuel; the emissions of those fuels are thus added up to calculate the total emissions of each sector.
Then, we allocate the sectoral emissions in the EB to 381 sectors in the IOT based on the energy transaction data in the IOT for different energy sources, with the assumption that each energy would be the same price in similar industries as the energy price is decided by the government. This was calculated by Eq. (2):
where CO2 i Emissions is the total amount of CO2 by energy i, xj is the sales volume for energy source i for industrial sub-sector j in the IOT, and
is the total output for the industrial sector by similar economic activity for energy source i.
To complete the EEIO table based on CO2 eq. data on six major GHG, such as CH4, N2O, HFCs, PFCs, and SF6 are supplemented from Korea’s GHG inventory, which covers all GHG from the country. CH4 and N2O emissions from energy sources are allocated to each sector in proportion to the energy consumption (CO2 emissions) allocated as above. However, given emissions from waste and lime production are unrelated to energy sources, they are allocated to each sector in proportion to the total output.
2.4. Carbon Footprint Analysis
To quantitatively analyze direct and indirect carbon footprint characteristics by industry, the analysis method was based on the 2017 high-resolution (~380 sectors) EEIO model [21]. Using the EEIO model [21], we maintain some sectors, such as natural rubber, bituminous coal, and raw sugar, whose total input values are all ‘0’, indicating their products were not domestically produced in 2017; the technical coefficients were set to ‘0’ [22]. Furthermore, the industrial activities of ‘manufacturing equipment repair’ and ‘industrial machinery and equipment repair’ were moved from the other manufacturing sector to the commercial sector. In addition, the industrial activities of ‘educational service (public, non-profit, industry)’, ‘medical and health care (public, nonprofit, industry)’, ‘social welfare service (public, nonprofit)’, and ‘cultural service (public)’ were calculated by changing from the public sector to the commercial sector [22]. As in the previous study, the original Korean IOT differentiates the electricity generation sector such as hydro power, thermal power, nuclear power, private power station, renewable energy, steam, chilled or hot water, and air conditioning supply [21]. We aggregate them into one electricity generation sector to reduce the EEIO model from 381 to 376 sectors [22].
The first analysis aims to quantify the GHG emissions generated by the Korean automobile industry in terms of direct and total emissions perspectives [30–32]. Direct GHG emissions (D), also known as production-based GHG emissions of the automobile sector can be calculated using Eq. (3):
where B and X are sectors GHG emissions intensity matrix and the total output of the automobile sector respectively.
Total GHG emission (T) of automobiles include both direct and indirect GHG emissions, also known as consumption-based GHG emissions triggered by final demand for automobiles (i.e. private consumption, government consumption, investment, and net exports) and was calculated by Eq. (4):
where A is the technology matrix derived from sectorial transactions where a column represents the direct requirement of a sector per dollar worth of output. I is the identity matrix, and f is the final demand matrix.
The second analysis characterizes Scope 1, Scope 2, and Scope 3 emissions for each economic sector. Scope 1 emission is the direct emissions from each sector to meet the final demand, Scope 2 emission refers to those from the use of electricity and steam; and Scope 3 represents a supply-chain or cradle-to-gate emissions that extend upstream of resource extraction [33–36]. Following a previous study [37], our estimation of Scope 1, Scope 2, and Scope 3 emissions is driven from the modification of the power series expansion of total GHG emissions. This was calculated as follows, according to Eq. (5):
where the right side of the Eq. (5) shows the ripple effect of GHG emissions in the supply chain by final demand [38, 39]. Bf is the GHG emissions generated by the final demand (Scope 1 emissions). BAf means the GHG emissions resulting from producing the intermediate unit input (such as resources, electricity, steam, heat, etc.) required to produce one unit of the final demand. Thus Scope 2 emissions (indirect emission from electricity and heat use) can be calculated by Eq. (6):
where Be is the direct GHG emissions intensity of the electricity generation sector, Aej is the input from the electricity generation sector to industry j, BeAejf is the GHG emissions from electricity and heat consumed in each industry to produce unit output (Scope 2). Thus, Scope 3 is total GHG emissions minus the sum of Scope 1 and Scope 2, was calculated as shown in Eq. (7):
The third analysis takes a deeper look into the supply chain emissions and identifies the main GHG emission contributors for each automobile product using Eq. (8):
where the ∧ symbol over B indicates diagonalization, f is 1 for the industry of interest and 0 for the rest of the industry. Since this Eq. (8), only considers the final demand for the industry of interest, it can be used to determine the emissions contribution of the supply chain and can provide insights to design climate policy and mitigation strategies.
The fourth analysis takes a deeper look into the final demand emissions and identifies major contributors for each automobile product using Eq. (9):
3. Results and Discussion
3.1. Direct and Total GHG Emissions from the Automobile Industry
The Korean IOT includes 7 automobile sectors among ~380 high resolution sectors from passenger cars, buses, trucks, special-use vehicles, trailers and containers, car engines, and automotive parts (Table 1, all sectors in Table S1 in Supplementary material). As shown in Table 1, direct and total GHG emission shows different characteristics of both the climate impact of automobile sectors in Korea and their relative significance. Direct GHG emissions from the 7 automobile sectors amount to 2,030 kt CO2 eq. or 0.3% of the national total GHG emission in 2017. Of the 2,030 kt CO2 eq. emissions, passenger car, automotive parts, and engine accounted for 40.5%, 31.4%, and 15.6%, respectively. The remaining sectors such as buses, trucks and special use vehicles trailers and containers show a very limited contribution to the total (12.5%). In comparison, the final demand for the Korean automobile industry induced 60,325 kt CO2 eq. throughout the upstream supply chain (cradle to gate), nearly 30 times as large as direct emissions. It contributes 8.4% to Korea’s total GHG emissions and 6.2% to Korean GDP. In the case of passenger cars, the total output was 72,089 Mil. USD and the GHG emissions were estimated at 823 kt CO2 eq. and the direct GHG emissions intensity was 11.4 kt CO2 eq./Mil. USD. On the other hand, the final demand for passenger cars was 72,073 Mil. USD and the total GHG emissions were 41,372 kt CO2 eq. Thus, the total GHG intensity of passenger cars was 574.0 kt CO2 eq./Mil. USD, which is a little higher than 459.7 kt/Mil USD for automobile production in the UK [28, 41]. There are also reports that the GHG emissions for a mid-sized car production is 7.5–11.5 ton CO2 eq. in China [11], and 5 tons for small-sized Renault Clio and 8.9 ton CO2 eq for overall passenger cars in the EU [66]. These differences are considered to be mainly attributed to the different GHG emission intensities of electricity generation and industrial technology in both countries.
The average carbon emission per car produced in Korea was 11.1 tons CO2 eq. This was calculated by dividing the total GHG emissions from passenger car production in Korea (41,372 kt CO2 eq.) by the number of passenger cars produced in Korea (3,735,399) [42]. According to the sustainability report [43] published by H motor company in 2018, the number of passenger cars produced domestically and overseas was 4,635,356 and the total amount of GHG generated during the production process was 24,906,019 ton CO2 eq., which means the average total GHG emission per car produced by H motor was 5.5 ton CO2 eq. [43]. According to data published by V company, the average total GHG emission per passenger car produced by the company was 6.5 ton CO2 eq. [44]. However, the 5.5–6.5 ton CO2 eq. per passenger car by the above two companies is ~ 50% lower than our study’s finding. This difference may be attributed to incomplete pLCA used by the two companies, which inevitably results in truncation errors.
3.2. Scope 1, 2, and 3 Emissions of the Automobile Industry
Table 2 shows the total GHG emissions by Scope from the automobile industry. The breakdown of the total GHG emission confirmed the Scope 1, Scope 2, and Scope 3 emissions, which were calculated using Eq. (3), Eq. (6), and Eq. (7). As shown in Table 2 for the automobile products, direct GHG emissions (Scope 1) account for only 3.0% (1.2 to 4.7%) of the total and Scope 2 emissions account for 2.9% (0.9 to 5.8%) of the total. The Corporate GHG guideline protocol mandatorily requests the amount of GHG emissions should include the direct emissions (Scope 1) and indirect emissions from the use of electricity and heat (Scope 2) of the corporate [45]. So, production GHG emissions (Scope 1 and Scope 2) in the Korean automobile industry can be translated to be 5.9% (4.2–7.7%) of the total. A clear fact that emerges from Table 2 is that Scope 3 emissions dominate the carbon footprint of Korean automobile products, contributing 94.1% (92.3–95.8%).
According to the sustainability report by H motor company in 2018, Scope 1, Scope 2, and Scope 3 emissions from automobile production were 886 kt CO2 eq. (3.6%), 1,937 kt CO2 eq. (7.8%), and 22,083 kt CO2 eq. (88.7%), excluding ‘uses of sold vehicles’ and ‘end-of-life treatment of sold vehicles’ sectors [43]. In addition, F motor company reported the GHG emissions in production Scope 1 (0.9%), Scope 2 (2.3%), and Scope 3 (96.8%), respectively [46]. These results with our findings in this study clearly confirm that major GHG emissions in the automobile industry are Scope 3, which comes from other industries in the upstream supply chain.
3.3. Supply Chain Contribution to the Automobile Carbon Footprint
The contribution of emissions from the supply chain associated with automobile production is presented in Fig. 3. We distinguish between direct electricity consumption (Scope 2) and indirect electricity consumption, embedded in the supply chain (Scope 3) to identify the electricity consumption pattern. For example, producing passenger cars requires primary metal products, and the electricity needed to make these primary metal products is then classified as indirect electricity consumption.
As shown in Fig. 3, the primary metal product sector made a significant contribution and accounted for 31.4–57.2% (average 40.9%) of the automobile industry’s total GHG emissions. The electricity consumption sector, including direct and indirect electricity generation (categories 1 and 2 combined), contributes 25.3–34.7% (average 32.5%) of the total GHG emissions. These two sectors accounted for 63.7–80.0% (average 70.5%) of the automobile industry’s carbon footprint. Meanwhile, other sectors like mining products, chemicals, non-metallic minerals products, processing metal products, and distribution contributed 1.4–1.6%, 2.5–6.8%, 1.0–1.4%, 0.8–2.4%, 3.4–5.7%, and 7.9–19.9%, respectively. Considering these results, reducing the automobile industry’s carbon footprint needs to be strategically prioritized based on the contribution and current situations.
For years, efforts have been made to lower vehicle weight by reducing primary metal contents [47–49]. According to H motor company, the steel composition of small cars has changed from 67.9% in 2004 [50] to 63% in 2010 [73]. These trends show that steel products used in the automobile have been gradually declining through many years of continuous effort, but they still account for a high proportion.
In the meantime, the Korean steel industry has already relatively high energy efficiency, but its potential for energy savings is very limited [51], Thus, to reduce the GHG emissions in producing steel products, the use of green and/or low-carbon electricity is crucial, which results in reducing GHG emissions in the automobile sector. Though the carbon intensity of Korea’s electricity grid has slowly improved, it was 0.4403 ton CO2/MWh in 2022 [26], which is higher compared to the OECD average of 0.3804 ton CO2/MWh in 2017 [52]. The current Korean energy grid GHG emission intensity and the detailed contribution analysis suggest that reducing GHG emissions in the Korean automobile sector would benefit most from reducing metal products and transitioning to a low-carbon electricity grid, but mitigation efforts in other areas of the supply chain, such as replacing primary metal products, are also very important.
3.4. Drivers of GHG Emissions in the Automobile Industry and Their Implication
The total GHG emissions driven by the final demand sector are summarized in Table 3. Consumption accounted for 28.3% of the final demand, amounting to 28,323 Mil. USD, while investment, which accounted for 25.4% of the final demand, was 25,456 Mil. USD. Meanwhile, exports and imports were 64,501 Mil. USD, 18,042 Mil. USD, respectively. Therefore, net exports (exports minus imports) amounted to 46,458 Mil. USD, accounting for 46.3% of the final demand. In particular, the passenger cars sector accounted for 68.6% of the total automobile industry. On the other hand, the GHG emissions caused by consumption was 16,334 kt CO2 eq., which accounted for 27.1% of the final demand, while the investment-induced GHG emissions were 14,796 kt CO2 eq. accounted for 24.5% of the final demand. Meanwhile, the GHG emissions driven by the exports and imports activities were 40,101 kt CO2 eq., and 10,906 kt CO2 eq., respectively; therefore, net export’s GHG emissions were 29,195 kt CO2 eq., accounting for 48.4% of the final demand. In the case of the passenger cars sector, consumption, investment, exports, and imports induced GHG emissions were 15,832 kt CO2 eq. (38.3%), 9,569 kt CO2 eq. (23.1%), 22,322 kt CO2 eq. (54.0%), and 6,352 kt CO2 eq. (−15.4%), respectively. In addition, Table 3 shows that among the total GHG emissions of the automobile sub-sectors, the amount of GHG induced by exports was the highest, and most of the total GHG emissions from the automobile industry were generated from passenger cars.
As Korea is global automobile exporter, it is meaningful to estimate the impact of global GHG mitigation effort on Korea automobile export. The potential adoption of CBT system by Europe and the US as mentioned in the introduction part [16–17] could significantly impact Korea’s economy, particularly through its automobile export [53]. Based on this study, GHG emissions from the production of a single automobile are measured at 11.1 ton CO2 eq. per passenger car and we assume that a CBT of per ton CO2 eq. for passenger cars is 75 USD (following the National Assembly Futures Institute publication) [54], so 830.7 USD per passenger car will be charged. Overall, a carbon tax will be expected to add about 4.1% of the price of vehicles produced in Korea. Because of this condition, prices of export products are predicted to increase, and countermeasures are highly needed in advance to strengthen climate competitiveness.
4. Conclusions
The Korean automobile industry plays a vital role in driving the economy, but its carbon footprint needs urgent attention. This research employed a holistic life cycle perspective to quantify Scope 1, 2, and 3 emissions, to identify the contributors and drivers of direct and total GHG emissions. This research found that the direct emissions (Scope 1) are low and indirect emissions (Scope 2 and Scope 3) in the supply chain are dominant. Especially, primary metal production and electricity generation create the biggest carbon burden. Thus, comprehensive GHG emission strategies such as reducing metal use and transitioning to clean energy are the key suggestions from this study for reducing GHG emissions in the automobile sector. Finally, while our EEIO model has been used for the automobile sector, it can be used, built upon and refined in other sectors by future carbon footprint studies in Korea and the world.
PDF Links
PubReader
Full text via DOI
Download Citation 
Supplement
Print 
