Has the Emissions Trading System in Korea (K-ETS) been effective in reducing their emissions? Who are the largest corporate CO2 emitters in South Korea and how are they affected by the system? In this blog post, I will analyze 2019 corporate carbon emissions data from South Korea’s Greenhouse Gas Inventory and Research Center in combination with the K-ETS CO2 permit allocation data provided by Newstapa. After reading this article, you’ll find out
- Have corporate emissions decreased since the launch of K-ETS system in 2015?
- Who are the top 10 corporate carbon emitters?
- Comparison between their actual carbon emissions and allocation of CO2 allowance (permits)
In 2015, South Korea launched Emissions Trading System (K-ETS) to reduce its greenhouse gas emissions by 24.4% below the 2017 level. It is a cap-and-trade system that puts a cap on carbon emissions on 685 companies by distributing CO2 allowances (permits) based on individual company’s emissions history. That is, if a company is carbon-intensive, it would receive more CO2 permits than those that are less carbon-intensive. If a company emits more CO2 than the allocated amount of allowances, it would have to purchase permits from other companies, which cost about $27.62 per tonne of CO2 equivalent in 2020 (World Bank Carbon Pricing Dashboard). For more information on K-ETS system, check out the latest report from the International Carbon Action Partnership (ICAP).
K-ETS system has received much criticism for being too generous with the free allocation of allowances. Newstapa, an independent journalism agency in South Korea, recently revealed that the carbon-intensive companies received more allowances than their actual emissions. This gave the companies no incentive to reduce carbon emissions. They even made an extra profit by selling the carbon allowances!
Newstapa’s investigation result was a shock for me. I couldn’t believe that a system that was designed to stimulate corporate climate action could be failing that hard. So I decided to dig into the data that were used for the investigation.
South Korea’s Greenhouse Gas Inventory and Research Center publishes corporate carbon emissions data annually. We will analyze data from 2015 until 2019. Newstapa has made the CO2 permit allocation data publicly available on their Data Portal. Both datasets are in Korean.
I merged the datasets by company name using the VLOOKUP function on Google Sheet. Then I applied the GOOGLETRANSLATE function to automatically translate the excel file. If you ever need to translate a spreadsheet, I highly recommend using this function!
The resulting dataset contains names of companies subjected to carbon pricing, their sectors, annual carbon emissions, and allowances that they received for free from 2015 until 2019. Before we begin our analysis, we will clean the dataset using the following code.
co2_raw <- read.csv("data/kr_corporate_co2.csv") co2 <- co2_raw %>% mutate(allowance_t = replace_na(allowance_t, "0")) %>% # Replace NAs in allowance with 0 mutate_at(vars(emissions_t, allowance_t), as.numeric) %>% # Change column types drop_na(emissions_t) #Drop NAs in emissions
Since the launch of K-ETS in 2015, corporate emissions increased by 5.7% (627 million tCO2eq). The increase was the highest in 2018 (close to 10%!). Visually, we can easily see that the emissions (reported at the end of the year) and the allowances (announced at the beginning of the year) are highly correlated. Since I won’t do a linear regression analysis that controls for omitted variables such as GDP here, I cannot say whether there is a causal relationship between the two. But I would guess that the emissions reductions could be higher if CO2 allowances were set to a lower level.
# Summarize corporate-level data by year co2_summary <- co2 %>% group_by(year) %>% summarize( total_emissions = sum(emissions_t)/1000000, total_allowances = sum(allowance_t)/1000000 ) %>% mutate(emissions_2015 = total_emissions[which(year==2015)], prc_diff = round((total_emissions-emissions_2015)/emissions_2015*100, 2)) # Calculate percentage difference relative to 2015 ggplot(co2_summary, aes(x=year)) + geom_bar(aes(y=total_allowances, fill=year), stat="identity", width=0.65) + geom_line(aes(y=total_emissions), color="red") + geom_point(aes(y=total_emissions), color="red") + ylim(0, 800) + theme_classic() + labs(y="in million tCO2eq", title = "Change in corporate CO2 emissions and allowances", subtitle = "Red line: annual emissions, blue bars: annual allowances") + theme(axis.title.x = element_blank(), legend.position="none")
POSCO, Korea South-East Power, Korea Southern Power Co.Ltd., Korea Western Power, Korea Midland Power, Hyundai Steel, Samsung Electronics, Ssangyong Cement, and S-Oil emitted the most carbon dioxide in South Korea. Five of them are electricity generation companies. (This is not surprising because less than 5% of the electricity comes from renewable sources).
What drew my attention to this table is the fact that the top emitter, POSCO and Ssangyoung Cement Industry received more allowance than their actual emissions. Is it because POSCO and Ssangyoung actually reduced carbon emissions or was the cap not high enough for these companies? This question remains ambiguous.
# Top 10 carbon-emitting companies co2_2019 <- co2 %>% filter(year==2019) %>% arrange(desc(emissions_t)) %>% mutate(emissions_mil.t = round(emissions_t/1000000, 2), allowance_mil.t = round(allowance_t/1000000, 2)) %>% select(-sector_kor, -energy_consumption_tj, -year, -allowance_t, -emissions_t, -sector_eng_big) head(co2_2019, 10) %>% kbl(caption="Top 10 carbon-intensive South Korean companies in 2019") %>% kable_classic()
While answering this question is also out of the scope of this blog, I will visually explore the emissions and allowance data for POSCO and Ssangyong Cement. The red line represents annual emissions and the blue line represents annual allowance granted to the company.
posco_co2 <- co2 %>% filter(name_eng %in% c("POSCO")) %>% # Subset POSCO data mutate(prc_diff = round((allowance_t - emissions_t)/emissions_t*100,2)) # Calculate difference between emissions and allowance in each year posco <- ggplot(posco_co2, aes(x=year)) + geom_line(aes(y=emissions_t/1000000), color="red") + geom_line(aes(y=allowance_t/1000000), color="blue") + theme_classic() + labs(title = "POSCO", y = "in million tCO2eq") ssangyong_co2 <- co2 %>% filter(name_eng == "Ssangyong Cement Industrial Co., Ltd.") %>% mutate(prc_diff = round((allowance_t - emissions_t)/emissions_t*100,2)) ssangyong <- ggplot(ssangyong_co2, aes(x=year)) + geom_line(aes(y=emissions_t/1000000), color="red") + geom_line(aes(y=allowance_t/1000000), color="blue") + theme_classic() + labs(title = "Ssangyong Cement", y = "in million tCO2eq") ggarrange(posco, ssangyong)
POSCO’s emissions have been drastically increased since 2016. Until 2018, the amount of allowance granted to this company was higher than their emissions up to about 14.22% (2017). Peculiarly, even though POSCO’s emissions increased in 2016-2017, POSCO was financially rewarded (POSCO could make profits from selling the surplus allowances).
Ssanyong’s emissions have been decreasing steadily since 2015. Since 2017, their emissions level decreased below the allowance by 3-4%.
I would love to continue exploring the relationship between corporate emissions and allowances. How can we measure the effectiveness of the K-ETS system on reducing corporate carbon emissions? How do we optimize the number of allowances? How can we prevent financially rewarding companies that did not decrease carbon emissions, like POSCO?
If you have any feedback or comments on this exploratory analysis or ideas for future research, I would like to hear from you!
This post was originally published on Uni Lee’s Conscious Table.