Emission Control Routes in Liner Shipping between Korea and Japan

Abstract

Maritime shipping is considered a major air pollutant that directly affects global warming and climate change. This study aims to design an emission control route (ECR) that can contribute to long-term initiatives along with emission prevention policies. Applying a scenario analysis, this paper has analyzed the impact of ECR implementation on air quality in the shipping route between Korea and Japan. In this study, the ECR is defined as eco-friendly shipping routes that recommend maintaining a specific level of speed while sailing. Based on the navigation trajectories of 55 O/Ds, which were extracted by the automatic identification system (AIS) data of 135 container ships for each port call during one year, the ECR domains were designated. To analyze the introduction impacts, this study employed a scenario analysis based on the current Vessel Speed Reduction Program (VSRP) recommendation of 12 knots or less. The results demonstrated that the introduction of ECR in the liner services between Korea and Japan would increase the average travel time by 10.47–16.98%. However, according to the results of the scenario analysis, the introduction of the ECR can reduce emissions by 25.65% to 39.73%. By suggesting the initial concept of ECR and analyzing the emission reduction effects of its implementation, this study provides useful insights for a balance between economic and environmental performance.

1. Introduction

The international shipping industry plays a central role in global economic exchange. Ships have been used as one of the most cost-effective modes of international transportation compared to other transportation options. However, it is considered to have relatively low environmental sustainability compared to other transportation industries in recent years, and the transition to an eco-friendly industry would take a long time.

In particular, CO₂ emissions from shipping operations are considered a major air pollutant that directly contributes to global warming and resulting climate change, making reducing emissions a high priority. Similarly, SO_X, NO_X, and CO, which are precursors to aerosols, have also been classified as air pollutants requiring long-term emission reduction, along with carbon dioxide (CO₂), which causes primary damage to the environment and human health, as well as secondary damage through chemical reactions.

Although shipping was not included in the Paris Agreement adopted at the 21st Conference of the Parties (COP21) to the United Nations Framework Convention on Climate Change, the International Maritime Organization (IMO) has set a quantitative environmental goal of reducing greenhouse gas emissions attributable to shipping by at least 50% by 2050 compared to 2018 while continuing to work on reducing the emissions of other air pollutants [1,2]. Additionally, various policies and regulations are being implemented not only by the IMO but also at regional and national levels to create an eco-friendly shipping industry at the international level. Current eco-friendly policies for the shipping industry have several directions and goals, such as regulatory aspects that monitor emissions, the introduction of eco-friendly equipment and fuels, the provision of economic incentives for compliance with recommended speeds, and market-based measures.

On the other side, traces of emissions from shipping tend to be concentrated along the routes that ships travel. Moreover, due to climatic factors, emissions can travel up to approximately 400 km from the point of emission and have harmful effects on the global environment and human health [3]. In particular, the proportion of emissions produced by the main engine is much higher than that emitted by the auxiliary engine and boiler, accounting for approximately 82% of the emissions in the cruise mode driven by the main engine [4]. Figure 1 shows the trajectory of the generation of CO₂, which directly affects global warming, among the emissions generated by ship operation [5]. Based on this, the initial concept of the ECR was designed, and its performance in the air quality sector was estimated for ECR introduction in this study. It is also significant in that it provides a new perspective and direction for eco-friendly policies.

Currently, most studies related to air pollution from the shipping industry focus on the evaluation of existing policies, the impact of air pollution from ships on port areas, and the competitiveness of green fuels. However, this study differs from others in that it suggests new policy concepts for the maritime industry by examining the scope and direction of new policies to create a green shipping industry. The scope of this study is on the maritime section of the liner between Korea and Japan. In the liner service between Korea and Japan, Japan ranks fourth among Korea’s major export destinations and third among Korea’s major import destinations as of 2022 [6]. In other words, South Korea and Japan are highly dependent on trade. Additionally, liner shipping is the largest emission source of air pollutants in the shipping industry [7].

Therefore, this study aims to define the initial concept of the ECR, an eco-friendly policy tool that can improve air quality. After reviewing the relevant literature, this study designates the ECR domains based on traffic trajectories. In the following section, an analysis of the emission reduction effect caused by shipping was carried out according to scenarios that set the recommended velocity level within the ECR.

2. Literature Review

2.1. Eco-Friendly Policies of IMO and Countries

The IMO is an organization that develops laws, regulations, and standards for the global shipping industry and promotes cooperation among governments in the development of shipping-related technologies to ensure the sustainability of the global shipping industry in economic, social, and environmental terms. As CO₂ emissions from the shipping industry contribute to global warming and the social and environmental damage caused by air pollutants such as SO_X, NO_X, and particulate matter (PM) in urban areas along port coasts are high on the international agenda, IMO has called for compliance with various regulations and recommendations to control air pollution from shipping operations. In particular, IMO has been addressing the issue of air pollution from the shipping industry since 2010 with Annex VI, an international treaty to control air pollution from ship operations under the MARPOL Convention. Annex VI contains emission control standards or compliance measures for the major air pollutants CO₂, SO_X, NO_X, and PM in Resolution A.963 (23), Regulation 13, and Regulation 14 [8,9,10].

Specifically, IMO encourages member states to establish emission control areas (ECAs) that include coastal waters near ports and allow primary and secondary compliance measures in waters within the ECA to reduce emissions of SO_X, NO_X, and PM from ships. In other words, the ECA is part of the IMO-established policy to progressively reduce emissions under MARPOL Annex VI. Although no specific emission standards have been established for PM, consideration is being given to reducing emissions associated with sulfur in the particulate formation process through the application of the marine fuel sulfur content requirements in Regulation 14 or through the introduction of additional technologies.

Similarly, a primary compliance option is to directly adopt the requirements for marine fuel sulfur content set by the relevant ECA, and a secondary compliance option is to install and operate equipment on board the vessel designed to reduce emissions when marine fuel with a relatively high sulfur content compared to low sulfur fuel oil is used in the ECA’s inland waters. For example, measures for reducing SO_X include the use of scrubbers or exhaust gas cleaning systems (EGCS), and for NO_X, exhaust gas recirculation, selective catalytic reduction (SCR), EGCS, scrubber, particulate filter, and silicon carbide diesel particulate filter can be used to marginally comply with the regulation requirement. Although it is not among the air pollutants subject to ECA regulation, in the case of CO₂, carbon capture and storage or carbon capture and utilization technology has been introduced on ships to reduce emissions and remove captured CO₂ deep underground, which can create added value by storing it or creating new products from it.

Countries in Asia currently implementing ECA policies include China, India, Malaysia, Singapore, South Korea, and Taiwan. In the case of Japan, a review of ECA implementation is underway. In Europe (Belgium, Estonia, Germany, Gibraltar, Italy, Portugal, Turkey), the Americas (Argentina, Bermuda, Brazil, California, Connecticut, Florida, Hawaii, Panama channel), Africa and the Middle East (Bahrain, Egypt, Kenya, Oman, Qatar, Saudi Arabia, South Africa, United Arab Emirates), and Oceania (Australia and New Zealand), emission regulations are implemented based on the ECA, with the intensity of regulations varying according to regional requirements. However, the regulation of the sulfur content of marine fuel for ships sailing in the open sea has been progressively tightened to a level of 4.5%S from May 2005 to December 2011 and 3.5%S from January 2012 to December 2019. The updated regulations are valid outside the ECA areas, and 0.1%S is being applied to vessels within ECAs after 2020 [11].

For the technical aspect, IMO also applies the mandatory Energy Efficiency Index for Existing Ships (EEXI), the Ship Energy Efficiency Management Plan (SEEMP), the Carbon Intensity Indicator (CII), the Recommended Energy Efficiency Operational Indicator (EEOI) for existing ships, and the mandatory Energy Efficiency Design Index (EEDI) for building new ships [12].

The green shipping policies currently implemented by IMO cover a wide range of areas aimed at reducing emissions. However, typical ECA policies are geographically limited to waters near ports or designate some waters within the baselines of territorial waters as ECAs. Vessel operators seeking to balance environmental sustainability with economic performance tend to increase their travel time in waters outside ECA boundaries and avoid entering ECAs due to the enforced regulation of ECAs [13]. Nevertheless, the ECR proposed in this study overcomes the geographic limitations of ECAs and allows vessel operators to participate voluntarily. It is also expected to help meet the standards applied by IMO for four eco-friendly indicators related to the operation of existing ships and to create synergy in achieving the IMO2050 ambition.

Various policies have also been taken at the national level to prevent air pollution caused by the shipping industry. To reduce emissions from shipping operations, South Korea and the United States have implemented a Vessel Speed Reduction Program (VSRP). The VSRP is a program that provides incentives for costs incurred by ships entering a port, such as fees for the use of port facilities and entry fees for ships if the ship enters the port at the average speed recommended in the speed reduction zone. Currently, the United States and South Korea are the two countries that have implemented the program, with the Port of Los Angeles and the Port of Long Beach in the United States and the Ports of Busan, Incheon, Yeosu-Gwangyang, and Ulsan in South Korea being the main target ports, with a recommended average speed for VSRP of 12 knots or less [14,15,16]. Like ECAs, VSRPs are geographically limited and regulate the entry and exit process. However, while VSRPs provide incentives for compliance with recommended vessel speeds on entry, there are no separate policies for exit. In contrast to the unilateral policy of VSRP, ECR recognizes the multidirectional nature of vessel operations and provides a motivating factor for vessel operators to voluntarily participate in creating a more eco-friendly environment for shipping.

2.2. Collaborated Invest and Market-Based Measures

As part of various policies to create an environment for a green shipping industry implemented at the international, national, and local levels, the construction of green shipping corridor (GSC) infrastructures is underway to achieve original zero emissions through public–private investment and the creation of collaborative consortia between industries. The Clydebank Declaration, signed at COP 26 by 22 countries with shipping and port industries, including the United States and the United Kingdom, has an initiative to gradually expand GSCs across the global trade network. Aggressive progress is being made, with a plan to establish 20 port infrastructures with green fuel supply chain functions by 2030 and to deploy at least 200 green ships in GSC. In the face of global climate change and social threats, the transition from conventional fossil fuels to green fuels for shipping is essential and a fundamental solution to ensure the sustainability of the shipping industry.

The range of green fuels, which is the most important factor for the establishment of GSC, can be classified as methanol, ammonia, biofuel, hydrogen, electricity, etc. Quantitative studies have been conducted to select suitable fuels according to the characteristics of each shipping sector and the behavior of shipping operations. The shipping segments that are likely to be developed as GSCs based on the feasibility of deployment are six container ship routes, six ferry routes, and one of each other route for bulk carriers, tankers, cruise ships, etc. Stakeholders participating in GSC construction include port authorities, ship owners/operators, fuel producers, governments, and financial institutions [17]. The GSC is a project to decarbonize the shipping industry, which is currently at an early stage due to its approach to building the infrastructure. Therefore, the expansion of the GSC network would take a long time. Indeed, building green infrastructure for the shipping industry in the global trade network requires a huge investment. Therefore, the creation of GSC requires policies that ease the financial burden on the industry to switch from fossil fuels to green fuels through green fuel pricing or direct support for the construction of green shipping and port infrastructure, as well as support for the additional costs associated with port use [18]. In particular, the liner shipping industry is cooperating internationally to establish four shipping segments as green shipping operators, as summarized in

Table 1. Green shipping corridor routes for liner shipping.

Appellation	Fuel	Led by	Target Timeline
Rotterdam–Singapore green and digital corridor	Methanol, ammonia	Public–private	latest by 2027
Rotterdam–West Coast Norway green corridor	Methanol	Public–private	latest by 2030
Shanghai–LA	Unknown	Industry or NGO	latest by 2030
SILK Alliance	Unknown	Industry or NGO	-

Source: [19].

[19].

There is also a market-based measure, the emissions trading scheme (ETS), based on shipping industry mechanisms. The main purpose of the ETS is to provide emission quotas for individual ships in the very carbon-intensive shipping industry and to allow for secondary trading at market-determined carbon prices. Countries that currently have the ETS include the United States, South Korea, Japan, New Zealand, and Switzerland. China plans to implement the ETS starting in 2021 [20]. If the carbon market is activated in the future and an international carbon market is formed by establishing an international standard for the ETS, a more effective and organic carbon market can likely be formed.

In the last five years, the size of the global carbon market has increased to 186 billion euros (2018), 240 billion EUR (2019), 288 billion EUR (2020), 762 billion EUR (2021), and 865 billion EUR (2022), with a growth rate of 365.05% in 2022 compared to 2018 [21].

As a result, GSC is in the early stages of introducing technology and infrastructure to achieve a green shipping network to achieve zero-emission initiatives, and it takes a long time to expand GSC into a network. Additionally, in the process of expanding the scope of the network, various factors such as the complex cooperation system among stakeholders, the feasibility of introducing GSC, and the financial situation of the port area where the expansion is required are complex; thus, introducing GSC into the overall shipping network and introducing green ships is a very significant investment in terms of time. Therefore, the introduction of ECR can help improve the operational efficiency of existing lines while building a GSC network. Furthermore, if a political basis for linking ECR with the ETS is established, it is expected that a high synergy effect can be achieved in preventing air pollution from ships.

3. Methodology

3.1. Concept of Emission Control Route

The ECR designed in this study is an initial concept for greening the shipping industry. The ECR is a fundamental design that can complement some of the categories in need of improvement that hinder the effectiveness of policies. The scope of ECA restrictions and VSRP recommendations is limited to the waters within the ports’ territorial sea baselines. However, as regulations are further tightened, ship operators may avoid ECA waters, even if it increases their travel distances. This could undermine the effectiveness of reducing emissions from vessel operations [22].

Additionally, the VSRP does not provide an incentive for outbound vessels, and vessels that pass through other ports’ RSZs are not eligible for the incentive; therefore, they do not maintain the specified speed. On the other hand, the scope of the slow speed recommendation is extended to the country’s EEZ or the open sea outside the baselines of territorial waters while designating the areas heavily frequented by container ships calling at ports between different countries as ECR areas.

The ECR concept proposed in this study is based on the characteristics of liner shipping that operates regularly on a fixed route as a policy target. The recommended speeds for container ships sailing in the ECR areas are based on the VSRP speed ranges in effect at the Port of LA and the Port of LB in the United States, and at the Ports of Busan, Incheon, Ulsan, Yeosu-Gwangyang, and Pyeongtaek-Dangjin in Korea. Figure 2 illustrates the concept of the ECR, and lag d, lag f, lag k, and leg o are designated as ECR domains in which recommendations about speed reduction are at a certain level.

Accordingly, the effects of reducing emissions were estimated by analyzing ECR control scenarios with criteria of 12 knots (Scenario 1), 11 knots (Scenario 2), and 10 knots (Scenario 3). If the prescribed speed per scenario and time interval were exceeded within the ECR ranges, the prescribed speed was applied, and if it was not exceeded, the vessel’s actual sailing speed per time interval was applied. The operational definition of ECR can be expressed as Equation (1), where ${Eff}_{p,q,s}$ is the emissions reduction effect of each type of emission “$p$” during the voyage of the container ship in each port call “$q$” of each of the introduced ECR scenario assumptions. Additionally, “$s$” is the assumed scenario level based on $s\in R (s=1,2,3)$.

$${Eff}_{p,q,s}=\left\{\sum {(emiss}_{p,q,non-ECR}-{emiss}_{p,q,s})\right\}$$

(1)

3.2. Case Study Specifications

3.2.1. Target Shipping Route Description

This study was conducted across a wide range of voyage trajectories of container ships used in liner shipping between Korea and Japan. South Korea’s exports to Japan were valued at 25,097,651,000 USD (2020), 30,061,806,000 USD (2021), and 30,606,278,000 USD (2022). In addition, South Korean imports from Japan were estimated at 46,023,036,000 USD (2020), 54,642,165,000 USD (2021), and 54,711,795,000 USD (2022). Furthermore, among South Korea’s top export trading partners, Japan ranked fifth in 2021 and fourth in 2022, and among South Korea’s top water ice trading partners, Japan ranked third in both 2022 and 2023 [6]. In other words, it can be determined that the trade dependence between South Korea and Japan is quite high. In particular, the main import and export goods between Korea and Japan are finished and semi-finished products that are traded via containerized sea transportation. The volume of liner trade handled between Korea and Japan is equivalent to 9,604,000 twenty-foot equivalent units (TEU) of Korean imports to Japan and 17,984,000 TEU of Japanese exports in 2022 [21].

3.2.2. Classification of the Target Port Calls

The historical data used in this study from AIS were collected from 1 July 2021 to 30 June 2022, a total of one year, through an API platform that provides maritime traffic information services for conducting GIS-based analyses. The total number of port calls is summarized in

Table 2. Number of liner trades between Korea and Japan.

No.	Two-Way Trade	Count	No.	One-Way Trade	Count
1	Port of Busan ⟷ Port of Yokohama	127	1	Port of Tokyo → Port of Ulsan	22
2	Port of Busan ⟷ MOJI KO	75	2	Hibiki Port → Port of Busan	13
3	Port of Busan ⟷ Port of Hiroshima	58	3	Port of Busan → Port of Yatsushiro	12
4	Port of Busan ⟷ Port of Kobe	58	4	Port of Busan → Port of Mishima-Kawanoe	6
5	Port of Busan ⟷ Port of Shimizu	55	5	Port of Busan → Port of Fushiki	5
6	Port of Busan ⟷ Port of Hakata	42	6	Port of Busan → Kashima	4
7	Port of Busan ⟷ Port of Akita	30	7	Port of Busan → Port of Hitachinaka	4
8	Port of Busan ⟷ Port of Hachinohe	28	8	Port of Busan → NAOETSU	3
9	Port of Busan ⟷ Sakata	24	9	Port of Ulsan → Port of Yokohama	3
10	Port of Busan ⟷ Port of Sendai	23	10	Port of Incheon → Hibiki Port	2
11	Port of Busan ⟷ Port of Osaka	20	11	MOJI KO → Port of Gwangyang	2
12	Port of Busan ⟷ Port of Tokuyama	18	12	Port of Chiba → Port of Ulsan	2
13	Port of Busan ⟷ Port of Nagoya	17	13	Port of Hakata → Port of Ulsan	2
14	Port of Busan ⟷ Port of Tokyo	16	14	Port of Nagoya → Port of Gwangyang	2
15	Port of Busan ⟷ Port of Sakaiminato	13	15	Port of Onahama → Port of Busan	2
16	Port of Busan ⟷ Port of Chiba	10	16	Port of Kobe → Port of Gwangyang	2
17	Port of Busan ⟷ Port of Yokkaichi	10	17	Port of Hamada → Port of Busan	2
18	Port of Busan ⟷ KANAZAWA	8	18	Port of Busan → Port of Fukuyama	1
19	Port of Busan ⟷ Port of Naha	6	19	Port of Busan → Port of Imabari	1
20	Port of Busan ⟷ Port of Tsuruga	5	20	Port of Gwangyang → Port of Tokyo	1
21	Port of Ulsan ⟷ Port of Kobe	4	21	Port of Gwangyang → Hibiki Port	1
22	Port of Ulsan ⟷ Port of Mizushima	3	22	Port of Gwangyang → Port of Hachinohe	1
23	Port of Busan ⟷ Port of Mizushima	2	23	Port of Ulsan → Port of Shimizu	1
-	-	-	24	Port of Yeosu → Port of Kobe	1
-	-	-	25	HAKODATE → Port of Busan	1
-	-	-	26	MOJI KO → Port of Ulsan	1
-	-	-	27	Port of Shimizu → Port of Ulsan	1
-	-	-	28	Port of Takamatsu → Port of Ulsan	1
-	-	-	29	Port of Yokohama → Port of Ulsan	1
-	-	-	30	WAKAYAMA KU → Port of Busan	1
-	-	-	31	Port of Kawasaki → Port of Yeosu	1
-	-	-	32	Port of Busan → Port of Takamatsu	1
	Total	652		Total	103

. In detail, for the collected data, error checking was performed to check the heterogeneity of the data for the parts where voyage information and trajectory confirmation were ambiguous due to AIS data transmitting and receiving problems in time interval units during the voyage. Accordingly, the data for port calls found to have any error were averaged based on the emissions for container ships of the same class for 20 port calls, which is excluded in Table 2.

3.3. Research Methods

3.3.1. Kernel Density Estimation

The container vessels freely navigate the waters between the port of origin and the port of destination, taking into account the shortest route and economic efficiency, to provide regular maritime services to the assigned port call. Although the route and activity characteristics for maritime transport prove to be complex, other container ships calling at the same ports also perform maritime transport along similar routes. Thus, characteristic trajectory patterns can be extracted when considering the overall traffic flow. In this study, we used the accumulated historical data from AIS generated by the port calls of container ships between Korea and Japan to comprehensively consider the temporal and spatial characteristics associated with the voyage trajectories of container ships and to perform ECR domain assignment based on kernel density estimation (KDE), a non-supervised method. ECR domains visualize spatial information about the segments that make up a ship route in the form of a geographic distribution.

In this study, we applied a KDE analysis to determine the spatial extent of ECR. KDE analysis is a well-established tool for analyzing traffic or transportation mode data with spatial features [23,24]. Additionally, KDE technology for geographic information systems (GIS) has become a leading method for determining whether phenomena occurring in space form clusters. Point-based detection is primarily in the form of traffic information that generates coordinates and travel times for specific passage points. In other words, the data containing the temporal and spatial information of the ship that has approached the port is expressed in terms of waypoints, and the function of KDE can be expressed according to Equation (2). In the equation, ${\left\{x_i,y_i\right\}}_{i=1}^n$ with $x_i,y_i\in R (i=1,2,3,\cdots ,n)$ represents the coordinates of the i-th waypoint, “$x_i-x$” and “$y_i-y$” represent the Euclidean distance between each waypoint “i”, and “h” represents the bandwidth [25].

$$\widehat{f_h}(x,y)=-\frac{1}{n{h}_x{h}_y}\sum _{i=1}^{n}K\left(\sqrt{{\left(\frac{x_i-x}{h_x}\right)}^2+{\left(\frac{y_i-y}{h_y}\right)}^2}\right)$$

(2)

In this study, the point-based data of container ships entering each port were converted into ship voyage trajectories based on the MMSI of each ship using the point-based analysis. Compared to the KDE for waypoints, the KDE for segments of trajectories can better extract waters with similar characteristics between trajectories of complex visualized maritime traffic and enable a more refined definition of ECR domains. As a result, a point-based KDE can be converted into a line-based KDE, which is represented by Equation (3). In the above formula, “$l_i$” represents the length of the line within the circular area, and “$w_i$” represents the weighting factor assigned by the KDE interpolation algorithm with $l_i,w_i\in R (i=1,2,3,\cdots ,n)$. Also, the area of the circular region indicates the radius to examine for each raster cell [26].

$$Line Density=\frac{\sum _{i=1}^n(l_i\times w_i)}{Area of Circle}$$

(3)

3.3.2. Evaluating Emissions

This study estimates emissions of each type of air pollutant for each scenario using a bottom-up approach based on parameters that can be calculated using primary information on vessel activity and operational information. The data used for the emissions estimation include the identification of each container ship during the port calls and the operational characteristics, such as the real-time position, speed, heading, and navigation status of the container vessels at a single port call, ME specifications, etc. The bottom-up approach based on data from AIS is suitable for detailed analysis of temporal and spatial variations in emissions compared to the top-down approach based on fuel. It is also possible to improve the estimation accuracy of emissions that vary with high-quality per time intervals depending on the dynamic features of the ship. Therefore, an AIS-based approach was used to quantify challenging emissions. However, this study found that the variability of sea conditions and meteorological changes affecting shipping have a relatively small impact on emissions attributable to ship operations. In addition, due to limited data availability, it was assumed that the influence of sea conditions and weather factors on ship operations is small [27]. For container ships, the functions of each engine type are categorized as follows: main engine (ME) for propulsion, auxiliary engine (AE) for power supply for life on board or operation of control systems, and auxiliary boiler (AB) for heating, hot water generation, and heating to maintain fuel liquidity. Furthermore, when the ship is in cruise mode, ME, AE, and AB are operated continuously [28]. Additionally, the ship’s operating mode was classified into 0 knots (berthing, stationary, and anchorage), 1–5 knots (maneuvering), and above 5 knots (cruising) depending on ship speed [4]. On this basis, the total amount of emissions generated by the operation of container ships in liner service between Korea and Japan during the period under study can be expressed by Equation (4), where $\sum {Emiss}_{p,q,r}$ is the total emission volume of air pollutants “$p$” for engine type “$r$” of each port call for liner shipping “$q$”; ${MEeg}_{p,q}$, ${AEeg}_{p,q}$, and ${ABeg}_{p,q}$ represent the amount of energy used to run the main engine [kW], auxiliary engine energy [kW], and auxiliary boiler energy [kW], respectively, about “$p$” and “$q$”; and ${EF}_{p,r}$ is emission factors [g] for “$p$” and “$r$”.

$$\sum {Emiss}_{p,q,r}=\sum \left({MEeg}_{p,q}·{EF}_{ME}+{AEeg}_{p,q}·{EF}_{AE}+{ABeg}_{p,q}·{EF}_{AB}\right)$$

(4)

Emission factors (EFs) based on energy units were used to estimate emissions based on the type of air pollutants attributable to ocean-going vessels. With the tightening of Tier 3 regulations related to NO_X and the requirement that bunker oil contains less than 0.5%S when sailing in waters outside the ECA, ships are finding alternative ways to comply while sailing in the open ocean or using bunker oil with less than 0.5%S. In this study, due to limited access to information on detailed compliance measures and available fuels for each container vessel in the collected data, ships were assumed to have used HFO 0.5%S for port calls between Korea and Japan during the data collection period. The EFs for the major air pollutants considered in this study are summarized in

Table 3. Emission factors for each air pollutant based on engine type and marine fuel oil.

Classification			Emission Factors (Unit: g/kWh)
Engine Type	IMO Regulations		Type of Air Pollutants
	SOX	NOX	CO2	CO	SOX	NOX
ME	HFO-0.5%S	Tier 3	600	1	2.1	2.5
AE	MD	-	661.4	1.1	2.2	13.9
AB	MD	-	970	0.2	16.5	2.1

Source: [29,30].

, which is reorganized based on the appropriate default values introduced in previous research [29,30].

The energy consumption per engine type, which is an input variable in Equation (4), is expressed in Equation (5) and calculated as the product of the distance traveled by the container ship at each port call and the energy consumed to operate the ME, AE, and AB [31]. During the analysis process, the total energy consumption of ME was calculated based on the dynamic information of the ship, including its distance and speed, recorded at a specific time during the voyage of each container ship, and the specific fuel-oil consumption (SFOC) calculated based on the maximum continuous rating (MCR), which is defined as the tested power by the manufacturers. The energy consumption for operating ME was calculated for each container ship by classifying them into high-speed diesel (HSD), medium-speed diesel (MSD), and low-speed diesel (SSD) based on the revolutions per minute (RPM) of the engine. The criteria established by RPM for categorizing MEs are HSD (${1200 \mathrm{rpm}<RPM}_{ME}$), MSD (${400 \mathrm{rpm}<RPM}_{ME}<1200 \mathrm{rpm}$), and SSD (${RPM}_{ME}<400 \mathrm{rpm}$). For the total energy consumption of AE and AB, the energy consumption of AE is 0.22 times that of ME due to the limited accessibility of specification information [32]. In addition, the ratio of energy consumption of AE and AB was calculated for AB by categorizing the classification of each container ship into a TEU [33], where $\sum {eg}_{q,r}$ is total consumed energy during each port call “$q$” of operation engine “$r$”, “$D$” is the transit distance for each track edge [Nm], “$V$” is the actual speed for each track edge [Knot], “${MCR}_{ME}$” is the maximum continuous rating of the ME [kW], “${LF}_{ME}$” is the load factor of each main engine [%], and “${SFOC}_{ME}$” is the specific fuel-oil consumption of each main engine [%].

$$\sum {eg}_{q,r}=\sum \left\{\frac{D}{V}\left[{MCR}_{ME}·{LF}_{ME}·{SFOC}_{ME}+{AEeg}_{p,q}+{ABeg}_{p,q}\right]\right\}$$

(5)

For SFOC, we categorized two-stroke engines and four-stroke engines according to how the ME works. Two-stroke engines have a relatively short combustion cycle to produce a stroke of power, and the functions within the combustion engine are concentrated compared to four-stroke engines. For this reason, the SFOC of a two-stroke engine is relatively lower than that of a four-stroke engine. SFOC is calculated with MCR as input, using Equations (6) and (7) as a function of the number of strokes [34], where “$s$” is the main engine type (two-stroke or four-stroke). Equation (6) is adopted when the main engine type is two-stroke; otherwise, Equation (7) is adopted.

$${ SFOC}_{deviation},s=0.0028·{MCR}^2-0.41·MCR+15$$

(6)

$${ SFOC}_{deviation},s=0.0036·{MCR}^2-0.58·MCR+23$$

(7)

The input variable of ${MEeg}_{p,q}$, the load factor (${LF}_{ME}$), corresponds to the propulsion power and can be expressed as a cubic number of the actual operating speed of a container ship and the design speed, which is the draft of the ship, according to a cubic relationship [35]. Equation (8) shows the relationship between ${LF}_{ME}$ and ${MEeg}_{p,q}$. The correction factor $k$ is 0.85 to account for the fact that ships do not use the full power of their engines during sea transport [36], where ${LF}_{ME}$ is the load factor for ${MEeg}_{p,q}$ [%], $V$ is the actual transit speed [Knot], $V_{Design}$ is the maximum designed speed [Knot], $P$ is the propulsive power for $V$ [kW], $P_{Max}$ is the maximum propulsive power for $V_{Design}$ [kW], and $k$ is 0.85, which indicates the fraction of installed power for $V_{Design}$.

$${LF}_{ME}=\frac{P}{P_{Max}}=k{\left(\frac{V}{V_{Design}}\right)}^3$$

(8)

4. Empirical Analysis

4.1. Designation of ECR

Japan does not currently have ECA, but it is considered a country that is likely to adopt ECA in the future [37]. Therefore, in this study, we assumed that Japan would adopt the ECA waters within the coastline that includes Hokkaido, Honshu, Shikoku, and Kyushu. Japan’s potential ECA was defined as the yellow line in Figure 3.

We used Q-GIS spatial data analysis software to analyze the designation of ECR areas and WGS 84 (World Geodetic System 84) to define the same spatial preferences in Q-GIS as the collected GIS data. First, we visualized the historical data from AIS by converting the dynamic attribute data with temporal and spatial changes from point (coordinate) form to line (trajectory) form for 135 container ships used for 755 port calls between Korea and Japan, defined as the red lines in Figure 3. In addition, the line representing South Korea’s territorial waters is colored blue. Second, considering the purpose of the study and the targeted geographic scope of the unique maritime traffic routes, the search radius was set to 30 km, and the frequency and pattern of maritime traffic were extracted in spatial form and represented by blue-colored shapes. Finally, a polygon with raster properties derived from the KDE analysis of the area with high navigation frequency was included, and a polygon in contact with Japan’s potential ECA was derived and set as the ECR domain. The results of the analysis steps and the ECR areas defined in this study are summarized in Figure 3.

4.2. Scenario Analysis

4.2.1. Air Pollution Prevention Effects of Each Scenario

A detailed analysis of the emissions estimation based on each ECR scenario was per-formed, combining ship activity data with ship-specific information that reflects the specific conditions per time interval. In detail, Figure 4 and Figure 5 show the estimates of ${\mathrm{C}\mathrm{O}}_2$, ${\mathrm{N}\mathrm{O}}_{\mathrm{X}}$, ${\mathrm{S}\mathrm{O}}_{\mathrm{X}}$, and $\mathrm{C}\mathrm{O}$ emissions for the non-ECR case, and scenarios 1–3 demonstrate the ECR broken down based on the port of origin and port of destination. In addition, the total reduction rates for the scenarios of each air pollutant based on the port call considered in this study were calculated, as shown in

Table 4. Emission reduction rates for each O/D.

No.	Origin–Destination	Reduction Rate	No.	Origin–Destination	Reduction Rate
1	Port of Busan → NAOETSU	−35.79%	29	Port of Gwangyang → Hibiki Port	−17.03%
2	Port of Kawasaki → Port of Yeosu	−31.52%	30	Port of Busan → Port of Hitachinaka	−16.89%
3	Port of Busan ⟷ Port of Tokuyama	−29.33%	31	Port of Busan ⟷ Port of Nagoya	−16.84%
4	Port of Busan ⟷ Port of Chiba	−28.87%	32	Port of Busan ⟷ Port of Yokohama	−16.72%
5	Port of Nagoya → Port of Gwangyang	−26.09%	33	Port of Busan ⟷ KANAZAWA	−16.60%
6	Port of Incheon → Hibiki Port	−25.55%	34	Port of Busan ⟷ Port of Tsuruga	−16.32%
7	Port of Gwangyang → Port of Tokyo	−25.40%	35	WAKAYAMA KU → Port of Busan	−15.90%
8	Port of Busan ⟷ Port of Hakata	−25.25%	36	Port of Onahama → Port of Busan	−15.47%
9	Port of Ulsan → Port of Yokohama	−24.48%	37	Port of Busan ⟷ Sakata	−15.17%
10	Port of Busan → Port of Fukuyama	−24.37%	38	Port of Busan ⟷ Port of Mizushima	−14.22%
11	Port of Ulsan ⟷ Port of Mizushima	−22.75%	39	MOJI KO → Port of Ulsan	−14.07%
12	Port of Busan ⟷ Port of Shimizu	−22.54%	40	Port of Busan ⟷ Port of Sakaiminato	−13.95%
13	Port of Busan ⟷ Port of Sendai	−22.19%	41	Port of Yeosu → Port of Kobe	−13.76%
14	Port of Busan → Port of Mishima-Kawanoe	−21.59%	42	Port of Busan ⟷ Port of Kobe	−12.26%
15	Port of Shimizu → Port of Ulsan	−21.57%	43	Port of Busan ⟷ Port of Hiroshima	−11.84%
16	Port of Hakata → Port of Ulsan	−21.43%	44	Port of Tokyo → Port of Ulsan	−10.67%
17	Port of Busan ⟷ MOJI KO	−20.60%	45	Port of Busan ⟷ Port of Hachinohe	−10.35%
18	Port of Busan → Kashima	−20.48%	46	Port of Ulsan → Port of Shimizu	−9.34%
19	Hibiki Port → Port of Busan	−20.07%	47	Port of Busan ⟷ Port of Tokyo	−8.72%
20	Port of Busan ⟷ Port of Osaka	−19.80%	48	Port of Busan ⟷ Port of Yokkaichi	−7.98%
21	Port of Yokohama → Port of Ulsan	−19.35%	49	Port of Busan ⟷ Port of Naha	−7.02%
22	Port of Busan → Port of Fushiki	−19.26%	50	HAKODATE → Port of Busan	−6.70%
23	Port of Kobe → Port of Gwangyang	−19.15%	51	Port of Hamada → Port of Busan	−6.62%
24	Port of Ulsan ⟷ Port of Kobe	−18.99%	52	Port of Busan → Port of Takamatsu	−5.49%
25	MOJI KO → Port of Gwangyang	−18.30%	53	Port of Chiba → Port of Ulsan	−4.81%
26	Port of Takamatsu → Port of Ulsan	−18.14%	54	Port of Busan ⟷ Port of Akita	−3.68%
27	Port of Busan → Port of Imabari	−17.99%	55	Port of Gwangyang → Port of Hachinohe	−2.78%
28	Port of Busan → Port of Yatsushiro	−17.95%	-	-	-

. There were 19 port calls that showed a reduction rate of 20% to less than 40%, 26 port calls that showed a reduction rate of 10% to less than 20%, and 10 port calls that showed a reduction rate of less than 10%. The emission reduction rate of the port calls which have more than 50 frequency represent −16.72% (Port of Busan ⟷ Port of Yokohama), −20.60% (Port of Busan ⟷ MOJI KO), −12.26% (Port of Busan ⟷ Port of Kobe), −11.84% (Port of Busan ⟷ Port of Hiroshima), and −22.54% (Port of Busan ⟷ Port of Shimizu), respectively, and the average overall emissions reduction rate for port calls was −17.27%. In other words, the emission reduction effect resulting from the introduction of ECR was found to have wide-ranging effects on overall individual port calls, regardless of the number of port calls between Korean and Japanese ports.

4.2.2. Policy Effects and Implications

Table 5. Emission reductions achieved by applying the ECR scenarios.

Unit: ton
Division		Setting up a Recommended Environment
		Non-ECR	Scenario 1	Scenario 2	Scenario 3
Air pollutants	CO2	1,996,608.89	1,484,288.75 (−25.66%)	1,340,514.56 (−32.86%)	1,203,161.55 (−39.74%)
	NOX	19,329.12	14,365.56 (−25.68%)	129,83.61 (−32.83%)	11,652.21 (−39.72%)
	SOX	8371.91	6226.45 (−25.63%)	5616.03 (−32.92%)	5043.87 (−39.74%)
	CO	3169.87	2357.65 (−25.62%)	2127.31 (−32.89%)	1910.28 (−39.73%)
ECA external average voyage time (Unit: h)		46.42	51.28 (10.47%)	54.20 (16.76%)	54.30 (16.98%)
Mean seeds (Unit: knot)		12.18	11.41 (−6.32%)	10.89 (−10.60%)	10.67 (−12.40%)

shows the results of the analysis for each port call unit, broken down based on the port of departure and port of arrival, as well as the results of the scenario analysis for the introduction of ECR at the scenario level. Based on the overall results, the average travel time for non-ECR was 55.68 h, resulting in estimated emissions of 1,996,608.89 $\mathrm{t}\mathrm{o}\mathrm{n}\mathrm{s}$ (${\mathrm{C}\mathrm{O}}_2$), 19,329.12 $\mathrm{t}\mathrm{o}\mathrm{n}\mathrm{s}$ (${\mathrm{N}\mathrm{O}}_{\mathrm{X}}$), 3169.87 $\mathrm{t}\mathrm{o}\mathrm{n}\mathrm{s}$ (${\mathrm{S}\mathrm{O}}_{\mathrm{X}}$), and 2422.69 $\mathrm{t}\mathrm{o}\mathrm{n}\mathrm{s}$ ($\mathrm{C}\mathrm{O}$) for each air pollutant. However, in Scenario 1, the trave time increased by 10.47%, while the total emissions decreased by approximately 25.65%. In Scenario 2, the travel time increased by 16.76% and the emission reduction decreased to 32.89%, while under Scenario 3, the travel time increased by 16.98% and the emissions were reduced by approximately 39.73%. On average, the emission reduction efficiency for the increase in voyage time was found to be 2.45% (Scenario 1), 1.96% (Scenario 2), and 2.34% (Scenario 3), respectively. These results show that container ships not operating at ECR levels are, on average, operating at speeds above 12 knots, but close to it. In other words, it is believed that the introduction of ECR in the liner trade between Korea and Japan could have a significant impact on air pollution control.

Figure 6 shows the results of analyzing the change rate of linear speed for non-ECR, Scenario 1, Scenario 2, and Scenario 3. Non-ECR ranges from 10.08 to 13.82 knots, Scenario 1 ranges from 9.88 to 12.77 knots, Scenario 2 ranges from 9.82 to 12.50 knots, and Scenario 3 ranges from 9.72 to 12.30 knots. Container ships could achieve higher fuel efficiency by maintaining a constant speed. From the above analysis, it can be concluded that the introduction of the ECR can help improve the fuel efficiency of ships by performing the secondary function of reducing variations in ship speed.

5. Conclusions and Discussion

The purpose of this study is not to estimate the environmental changes in the shipping industry and the effectiveness of the current eco-friendly shipping policy. Instead, we aimed to review the current eco-friendly shipping policy in general and introduce the initial concept of ECR to explore a different direction of policy to prevent air pollution from ship operation. In addition to contributing to meeting various standards related to ship operations at IMO in the long-term for the creation of a GSC network, the ECR can serve as a means to improve the spatial coverage constraints of the RSZ for the implementation of the VSRP and to effectively create economic and environmental synergies in conjunction with the ETS in the future. Therefore, the ECR is a practical concept that can play a supporting and complementary role in current green shipping policies.

We have provided a conceptual definition of ECR and demonstrated the effectiveness of reducing emissions through a step-by-step scenario analysis of the ECR concept. In other words, we demonstrated the effectiveness of air pollution prevention for ECR, a new concept that can reduce ship-related air pollution from a new perspective and provide insights into new policy directions and options. Specifically, the ECR approach proposed in this study could lead to significant environmental improvements given the nature of liner shipping, which is closely linked to global supply chains, and its contribution to air pollution. It will also create a practical motivation for ship operators to take social responsibility and strive for a balance between economic and environmental performance. In other words, the ECR is considered feasible, given the institutional conditions that support air pollution prevention, precautionary measures through direct regulation, and the economic performance of shipping companies.

However, there are some prerequisites for achieving such results. First, an international organization such as the IMO is required, which could play a role in bringing together the various stakeholders involved in implementing ECRs and direct participation by states on a reciprocal basis. This will facilitate environmental policy cooperation among countries directly involved in the implementation of ECRs and play a systematic facilitation role in reviewing effective legislation, institution building, and specific detailed policy elements. Second, the effective use of the ECR policy instrument would benefit from expanding the scope of the existing vessel tracking system, which controls vessel traffic within a country’s EEZ, to include liner shipping and establish an integrated system for operations in selected waters encompassing ECR domains. Follow-up activities that can support the ECR implementation are also needed.

Developing strategies to create a green environment for the shipping industry is the best way to implement environmental stewardship, including the economic and social functions of the shipping industry on a global scale. The development of strategies and measures to further reduce air pollution as the shipping industry fully decarbonizes until it is essentially zero-emission in the long term should be continuously explored, such as the ECR and intermodal transport chains with short sea shipping (SSS), which is a transport policy concept [38]. In addition, it is necessary to conduct more detailed studies on the multifaceted performance evaluation of the ECR implementation, which should further improve the environmental sustainability of the shipping industry in the global logistics industry. In other aspects, detailed research regarding whether shipping companies and ship owners could accept the recommended speeds of the ECR in terms of traffic conditions and potential economical profits such as ETS is needed.