Analysis of Greenhouse Gas Emissions from Ships Visiting European Ports

Abstract

This study analyzes greenhouse gas emissions from ships visiting European ports between 2020 and 2023, utilizing data from the EU Monitoring, Reporting, and Verification (EU-MRV) system. It examines the impact of the FuelEU Maritime Regulation on four types of ships during this period. It discusses updates to MARPOL Annex VI, including the Global Fuel Standard (GFS) designed to reduce emissions. A line contour method is employed to estimate emissions, focusing on tankers, bulk carriers, general cargo ships, and container ships while adhering to European regulations. This method models operational variables such as deadweight and ship age to categorize vessels based on their energy efficiency. For ships visiting European ports in 2023, the FuelEU regulation limit is set at ${GHG}_{limit,EU2020}$ = 91.16 gCO_2eq/MJ, indicating that general cargo ships are the most energy-efficient type, while bulk carriers are the least efficient. According to IMO regulations, the limit is ${GHG}_{limit,IMO2008}$ = 93.3 gCO_2eq/MJ, meaning all ships fully comply with their respective limits. The study utilizes real data, and the contour approach has played a crucial role in evaluating greenhouse gas emissions, representing a significant advancement in the methodology, modeling, and analysis of a ship’s energy efficiency.

1. Introduction

Shipping accounts for 80% of global trade and is responsible for approximately 25% of greenhouse gas emissions. The maritime community is working towards ambitious targets outlined in the Revised IMO strategy, which aims to significantly reduce emissions from shipping and achieve net-zero levels by 2050.

DNV [1] compiled and summarized 42 emission reduction measures, with each measure categorized into one of eight categories: Energy Consumers, Energy Harvesting, Propulsion and Hull, Machinery, Operational, Fuel, Other, and Energy Converters. The characteristics of these measures are detailed in terms of various metrics, including emission reduction potential, capital (CAPEX) and operational (OPEX) expenditures, technological maturity, infrastructure availability, and their impact on three indexes: the Design Index, Operational Index, and Fuel-based Index. These indexes encompass the EEDI (Energy Efficiency Design Index), EEXI (Energy Efficiency Index for Existing Ships), CII (Carbon Intensity Index), ETS (Emissions Trading Scheme), FuelEU Maritime Regulation, and Greenhouse Gas Fuel Index (GFI).

One of the rapidly developing areas in this context is the use of alternative fuels.

Table 1. IMO and EU legislation on alternative fuels.

IMO		EU
Carbon Intensity Index (CII)	GHG Fuel Index (GFI)	FuelEU Maritime Regulation (FEUM)	Emission Trading System (ETS)
came into effect on 1 January 2023	will come into force by March 2027	came into force on 1 January 2025	came into force on 1 January 2024
MEPC.352(78) [2]; MEPC.353(78) [3]; MEPC.338(76) [4]; MEPC.354(78) [5]; MEPC.355(78) [6]; MEPC.400(83) [7].	Circular letter No. 5005 [8] Expected to be formally adopted in October 2025.	Regulation (EU) 2023/1805 [9].	Directive (EU) 2003/87/EC [10]; Directive (EU) 2023/959 [11].

presents the results of the efforts made by the International Maritime Organization (IMO) and the European Union (EU) to implement measures related to the adoption of alternative fuels and the reduction of greenhouse gas emissions from shipping.

Regulation 28 of MARPOL Annex VI establishes the Operational CII for vessels with a GT of 5000 or more. This regulation requires that vessels calculate their achieved operational CII annually and compare it to the Required Annual Operational CII to determine their operational carbon intensity rating, which ranges from A to E. If a vessel receives a D or E rating, it is necessary to develop and implement a corrective action plan as part of the Ship Energy Efficiency Management Plan (SEEMP) to improve carbon intensity performance. At the most recent MEPC session, it was decided that the reduction factors for the period following 2027 would increase by 2.625% each year, ultimately reaching 21.5% by 2030 [7].

Beginning in January 2024, the EU Emissions Trading System (ETS) will regulate CO₂ emissions from all ships exceeding 5000 GT that dock at European Union ports. The coverage of emissions will be introduced gradually: 40% in 2024, 70% in 2025, and reaching a full coverage of 100% by 2026. Additionally, emissions of methane and nitrous oxide will be incorporated into the regulation from 2026.

The EU ETS functions on a “cap and trade” model, requiring shipping companies to buy European Union Allowance (EUA) for the emissions they report. Alternatively, they can lower their emissions to stay within their allocated allowances. Emissions from voyages taking place entirely within the EU are fully covered at 100%. For voyages that begin or end outside the EU but make a stop at a port in the EEA, only 50% of the emissions are accounted for. The cap for shipping emissions within the EU ETS for the year 2024 is set at 1,386,051,745 allowances [12]. This cap will decrease progressively over time to align with the EU’s climate goals, which include a target of a 55% reduction in greenhouse gas emissions by 2030 compared to 1990 levels. Each EUA permits the emission of one metric ton of CO₂ or an equivalent amount of other greenhouse gases. The EU carbon market started in 2025 with a price of EUR 71.52, reached a peak of over EUR 81 in mid-February, and then fell below EUR 60 in early April. The expected average cost for the year is approximately EUR 75, reflecting a 15% increase compared to 2024 [13].

Von Malmborg [14] employs the Advocacy Coalition Framework (ACF) as a theoretical lens to analyze the policymaking process related to the FEUM. The FEUM includes several important provisions:

• It requires progressive reductions in greenhouse gas (GHG) intensity, aiming for an 80% decrease by 2050.
• By 2034, a 2% sub-quota for renewable fuels of non-biological origin will be implemented based on market evaluations.
• Ships docking in EU ports must an utilize onshore power supply while berthed, except in cases where zero-emission technologies are employed.

To meet the goals of the FEUM, numerous European nations are exploring and investigating Power-to-X (PtX) technologies. This refers to the process of converting electrical energy into another form, denoted as “X.” For example, electrical power can be transformed via electrolysis to generate hydrogen, which can be utilized independently or combined with nitrogen (N) or carbon (C) sourced from CO₂ to create new fuels, including ammonia, methanol, and methane [15].

A recent study [16] conducted a forward-looking life cycle assessment (LCA) to evaluate the environmental impacts of e-ammonia, e-methanol, e-Fischer Tropsch (FT) diesel, and e-liquefied natural gas (e-LNG) in the context of maritime applications in Europe. The findings revealed that e-ammonia, e-FT diesel, and e-methanol can meet the FuelEU Maritime targets for 2050, whereas e-LNG and ship-based carbon capture systems (SBCCSs) do not meet these standards.

Oh, et al. [17] assess the GHG emissions from shipping with onboard carbon capture (OCC) and demonstrate that the emissions reduction achieved by OCC on the main engine could enable fossil-fueled ships to meet the GHG intensity targets established by FEUM until 2044. To maintain compliance in subsequent years, OCC and storage systems will also need to address emissions from the auxiliary boiler and generator, improve capture rates, and adopt strategies to minimize upstream emissions.

The FuelEU Maritime regulation, along with the CII rules and the EU ETS, has been integrated into an analysis aimed at minimizing costs through speed optimization for container ships operating on the Asia–Europe route, specifically those with a capacity of 13,000 to 14,000 TEU [18]. The results indicate that the optimal speeds differ across various segments, depending on the strictness of these regulations. In regions of the European Economic Area (EEA) that are either fully or partially regulated, a slow steaming approach is preferred to decrease fuel usage, lower greenhouse gas emissions, and reduce compliance expenses. Conversely, in areas outside the EEA where these regulations do not apply, higher speeds are seen as the most effective way to guarantee timely deliveries.

Lastly, a study on Swedish maritime transport [19] explores various scenarios involving multiple factors: fuel costs, infrastructure expenses, prices for EU EUA, the FuelEU Maritime regulation, and the impacts on ships, as well as discount rates of 3%, 5%, and 10%. The main findings of the study are as follows:

• Low price signals for EU Allowances result in limited emissions reductions and could even lead to increased emissions.
• Reducing the EU ETS inclusion threshold from 5000 GT to 400 GT could lead to an additional 7–11% reduction in CO₂eq emissions.
• The FuelEU Maritime regulation is expected to have minimal short-term effects until around 2030.

At the MEPC 83 session (April 2025), the regulatory text for amendments to MARPOL Annex VI was finalized and ‘approved’ for circulation to the parties involved, in anticipation of their adoption at the second Extraordinary Session of the MEPC (MEPC/ES 2) scheduled for October 2025 [20]. If these amendments are adopted, they will come into effect on 1 March 2027. The amendments introduce a new Chapter 5 “Regulations on the IMO Net-Zero Framework” to Annex VI of MARPOL. A term “Attained annual GHG fuel intensity (attained annual GFI)” is defined. The physical meaning of the term is analogous to that of FuelEU.

The following sections of the study analyze the impact of the FuelEU Maritime Regulation on four types of vessels that frequently visited European ports over four years. It uses relevant emission data sourced from the EU MRV database. Furthermore, there is a distinct discussion regarding the recent updates to MARPOL Annex VI, which include the introduction of a Global Fuel Standard (GFS) by the International Maritime Organization (IMO) designed to mitigate greenhouse gas emissions. This section also compares the GFS with the existing FuelEU Regulation.

The contour line approach is employed to evaluate greenhouse gas emissions from ships arriving at European ports. GHG emissions are calculated according to the methodology described in Annex I of the European regulation for four vessel types: tankers, bulk carriers, general cargo ships, and container ships. This calculation depends on data from the EU Monitoring, Reporting, and Verification system and identifies the spatial variations of operational parameters such as ship age, DW, and GHG emission levels. GHG emissions are deemed unacceptable whenever the operational conditions fall within the unacceptable contour region.

The line contour methodology initially involves developing a model that characterizes the combined distribution of operational factors, specifically deadweight and ship age. This joint distribution is subsequently utilized to identify a set of operational traits sharing the same joint exceedance probability, which defines a contour. Lastly, by applying the GHG emission limit, a contour is created and used to identify extreme responses for all ships; those within this contour are classified as energy efficient.

2. GHG Emissions Requirements of the EU and IMO

2.1. GHG Intensity According to FuelEU

The methodology for the estimation of GHG intensity is outlined in Annex I [9]:

$${GHG}_{intencity}\left[{\displaystyle \frac{g{CO}_{2eq}}{MJ}}\right]=f_{wind}\times \left(WtT+TtW\right)$$

(1)

where $f_{wind}$ denotes a reward factor for wind-assisted propulsion, $WtT$ represents well-to-tank GHG emissions, and $TtW$ refers to tank-to-wake GHG emissions.

$$WtT={\displaystyle \frac{{\sum }_i^{n\mathrm{ }fuel}{M}_i\times {CO}_{2eq\mathrm{ }WtT,i}\times {LCV}_i+{\sum }_k^c{E}_k\times {CO}_{2eq\mathrm{ }electricity,\mathrm{ }k}}{{\sum }_i^{n\mathrm{ }fuel}{M}_i\times {LCV}_i\times {RWD}_i+\mathrm{ }{\sum }_k^c{E}_k}}$$

(2)

$$TtW={\displaystyle \frac{{\sum }_i^{n\mathrm{ }fuel}{\sum }_j^{m\mathrm{ }engine}{M}_{i,j}\times \left[\left(1-\frac{1}{100}{C}_{slip,\mathrm{ }j}\right)\times {CO}_{2eq\mathrm{ }TtW,i,j}+\left(\frac{1}{100}{C}_{slip,j}\times {CO}_{2eq\mathrm{ }TtW,i,j}\right)\right]}{{\sum }_i^{n\mathrm{ }fuel}{M}_i\times {LCV}_i\times {RWD}_i+\mathrm{ }{\sum }_k^c{E}_k}}$$

(3)

where $i$ represents the different fuel types used during the reporting period, $j$ identifies the various fuel consumers, $k$ refers to the connection points for onshore power supply (OPS), $n$ denotes the total number of fuel types delivered to the ship within the reporting period, $c$ indicates the total number of OPS connection points, and $m$ signifies the total number of fuel consumer units. The mass of the fuel type $i$ consumed by the fuel consumer unit $j$ is represented as $M_{ij}$ [gFuel]. The amount of electricity supplied to the ship via the OPS connection point $k$ is noted as $E_k$ [MJ]. Additionally, ${CO}_{2eqWtT,i}$ stands for the WtT GHG emission factor of fuel type $i$ [gCO_2eq/MJ], while ${{CO}_{2eq}}_{electricity,k}$ refers to the WtT GHG emission factor for the electricity provided to the ship at berth at connection point $k$ [gCO_2eq/MJ]. The lower calorific value of fuel $i$ is denoted as ${LCV}_i$ [MJ/gFuel]. The reward, ${RWD}_i$, is set at 2 for fuels of non-biological origin used from 1 January 2025 to 31 December 2033. For fuel uses at other times, ${RWD}_i$ = 1. Additionally, $C_{slipj}$ represents the non-combusted fuel coefficient as a percentage of the total mass of fuel $i$ consumed by fuel consumer unit $j$ [%]. The $C_{slip}$ factor accounts for both fugitive and slipped emissions.

$${CO}_{2eq,TtW,i,j}=\mathrm{ }{\left(\begin{array}{c}{C_{fCO}}_{2,i,j}\times {GWP}_{co2}+{C_{fCH}}_{4i,j}\times {GWP}_{CH4}+\\ {C_{fN2O}}_{i,j}\times {GWP}_{N2O}\end{array}\right)}_i$$

(4)

$${{CO}_{2eq,TtWslip}}_{i,j}=\mathrm{ }{\left(\begin{array}{c}{C_{sfCO}}_{2,i,j}\times {GWP}_{co2}+{C_{sfCH}}_{4,j}\times {GWP}_{CH4}+\\ {C_{sfN2O}}_j\times {GWP}_{N2O}\end{array}\right)}_i$$

(5)

where ${{CO}_{2eq,TtWslip}}_{i,j}$ represents TtW ${CO}_2$ equivalent emissions from slipped fuel type $i$ for the fuel consumer unit $j$ [gCO_2eq/gFuel]. The variables ${C_{sfCO}}_{2i,j}$, ${C_{sfCH}}_{4i,j}$, and ${C_{sfN2O}}_{i,j}$ indicate the TtW GHG emission factors associated with slipped fuel $i$ for the fuel consumer unit $j$ [gGHG/gFuel]. Specifically, the following values are used: $C_{sfCO2}=0$, $C_{sfN20}=0$, $C_{sfCH4}=1$. Additionally, ${GWP}_{CO2}$, ${GWP}_{CH4}$, and ${GWP}_{N20}$ represent the Global Warming Potentials of ${CO}_2$, ${CH}_4$, and $N_{2}O$, respectively, over a 100-year period.

The current study utilizes the values from the Fifth Assessment Report (AR5) according to the EMSA Recommendations. The specific values are ${{GWP}_{CO}}_2=1$, ${{GWP}_{CH}}_4=28$, ${GWP}_{N20}=265$ [21].

According to Article 4 (2) [9], a company is required to pay a FuelEU penalty for any of its ships that have a GHG intensity compliance deficit. If a ship fails to meet compliance standards for two or more reporting periods in a row, the penalty will increase. This increase is calculated by multiplying the penalty amount by $1+(n-1)⁄10$, where $n$ represents the total number of consecutive reporting periods in which the company has incurred a FuelEU penalty for that ship. The compliance balance ($CB,\left[g{\mathrm{CO}}_{2\mathrm{eq}}\right]$) for a ship’s GHG intensity is calculated as follows:

$$CB=\left({GHGIE}_{target}-{GHGIE}_{actual}\right)\times \left[{\sum }_i^{nfuel}{M}_i\times {LCV}_i+{\sum }_k^c{E}_k\right]$$

(6)

where ${GHGIE}_{target}$ refers to the GHG intensity limit of the energy used on board a ship, while ${GHGIE}_{actual}$ is the yearly average of the GHG intensity calculated for the relevant reporting period.

The calculation of FuelEU penalties (in EUR) will be conducted as follows:

$$FuelEU\mathrm{ }penalty={\displaystyle \frac{\left|CB\right|}{{GHGIE}_{actual}\times \mathrm{41,000}}}\times 2400$$

(7)

where 41,000 is the equivalent of 1 metric ton of VLSFO in MJ, and EUR 2400 is paid for an equivalent metric ton of VLSFO.

2.2. Attained Annual GFI According to IMO

To achieve the goals outlined in the revised IMO 2023 Greenhouse Gas Reduction Strategy, a new Chapter 5, titled “IMO Net-Zero Framework,” has been pre-approved for inclusion in MARPOL Annex VI [8]. Beginning at the end of the calendar year 2028, and for each calendar year thereafter, every ship is required to calculate its attained annual GHG fuel intensity (attained annual GFI) for the previous year, covering the 12 months from January 1 to December 31. The attained annual GFI (${GFI}_{attained}$) for a ship will be determined by

$${GFI}_{attained}={\displaystyle \frac{\sum _{j=1}^J{EI}_j\times {Energy}_j}{{Energy}_{total}}}$$

(8)

using the following parameters: $j$ represents the different types of fuel used, with $J$ indicating the total number of fuels reported in the IMO Ship Fuel Oil Consumption Database [22]. ${EI}_j$ is the GHG intensity of a fuel type $j$, measured on a well-to-wake basis in gCO_2eq/MJ. ${Energy}_j$ denotes the energy consumption of the fuel type $j$ by the ship during the reporting period expressed in MJ. ${Energy}_{total}$, also in MJ, refers to the total energy consumed by the ship in the same period.

This target annual GFI, ${GFI}_T$, is calculated as follows:

$${GFI}_T=\left(1-{\displaystyle \frac{Z_T}{100}}\right)\times {GFI}_{2008}$$

(9)

where $Z_T$ refers to the percentage reduction factor that will be applied annually to the reference value of 93.3 gCO_2eq/MJ, which was established for the year 2008.

The rules define reduction factors for two reduction trajectories: base target and direct compliance target (Figure 1). The assessment of compliance with these rules is based on the following:

$$GFI\mathrm{ }compliance\mathrm{ }balance\left({CO}_{2}eq\right)=\left(\begin{array}{c}DirectcompliancetargetannualGFI-\\ AttainedannualGFI\end{array}\right)\times {Energy}_{total}$$

(10)

If the GFI compliance balance is zero or more, the ship will be in full compliance and will be eligible to receive surplus units according to the regulations set by the IMO. Conversely, if the GFI compliance balance falls below zero, the ship must evaluate its compliance shortfall using two defined tiers.

$$Tier1\mathrm{ }compliancedeficit=\left(\begin{array}{c}DirectcompliancetargetannualGFI-\\ AttainedannualGFI\end{array}\right)\times {Energy}_{total}$$

(11)

When the $attained annual GFI$ exceeds the base target, then

$$Tier\mathrm{ }1\mathrm{ }compliance\mathrm{ }deficit=\left(\begin{array}{c}DirectcompliancetargetannualGFI-\\ BasetargetannualGFI\end{array}\right)\times {Energy}_{total}$$

(12)

and

$$Tier2\mathrm{ }compliancedeficit=\left(\begin{array}{c}BasetargetannualGFI-\\ AttainedannualGFI\end{array}\right)\times {Energy}_{total}$$

(13)

A ship can address its shortfall in Tier 1 compliance by purchasing remedial units through contributions to the IMO Net-Zero Fund for GHG emission pricing. To offset any shortfall in Tier 2 compliance, additional options are available, such as utilizing surplus credits from previous years. For the reporting periods from 2028 to 2030, the initial price for a Tier 1 remedial unit will be set at USD 100 per metric ton of CO₂eq on a WtW basis, while the initial price for a Tier 2 remedial unit will be USD 380 per metric ton of CO₂eq on a WtW basis.

The IMO defines technologies, fuels, and energy sources that result in zero or near-zero GHG emissions (referred to as ZNZs). The Greenhouse Gas Fuel Index (GFI) threshold for these ZNZs will be set at a maximum of 19.0 gCO_2eq/MJ for an initial period that will last until 31 December 2034 (see the green line in Figure 1 (right)). Starting 1 January 2035, this threshold will be reduced to a maximum of 14.0 gCO_2eq/MJ, according to the guidelines established by [23].

The GHG emissions requirements of the EU and IMO pursue the same goal, but at different speeds and with additional conditions. (Figure 1).

2.3. Comparisons of EU and IMO Requirements

Figure 1 illustrates the technical differences between the two sets of requirements in the rate of reduction of greenhouse gas emissions necessary to achieve the same goal. The figure highlights both the changes in the reduction factors (left) and the absolute values (right), which are based on differing initial levels. The base level for FuelEU is 91.16 gCO_2eq/MJ from 2020, and for GFI, the base is 93.30 gCO_2eq/MJ from 2008.

Following the announcement of the adopted texts in April of this year, several comparative analyses were published.

Bishnoi and Thakur [24] announce an overlap between FuelEU and IMO regulations, potentially leading to dual penalties. Until harmonization is established, vessels operating in the EU could be subject to compliance requirements from both FuelEU and the IMO, without assurance of mutual recognition. They draw attention to the fact that Tier 1 deficits must be addressed by paying penalties directly. The IMO Net-Zero Framework explicitly disallows pooling and trading of Tier 1 credits, which limits the flexibility for partially non-compliant ships.

Jallal [25] points out strategic differences between the two decarbonization regimes in terms of enforcement methods, economic consequences, and political context.

The analysis provided in [26] outlines the fundamental components of the IMO Net-Zero Framework and examines its commercial impacts, particularly regarding projections of compliance costs for major fuel types over time. Additional costs (other than fuel purchases) have been assessed under FuelEU (Penalty, Surplus, and EU ETS Costs) and IMO (Remedial Units (RU) 1, RU2, and Surplus). The analysis examines two scenarios: one where both IMO and EU regulations coexist and another where only IMO regulations are in effect.

In 2028, common maritime fuels such as HFO, LFO, and MDO/MGO will experience a modest rise in OPEX because of the combined impacts of FuelEU Maritime and the EU ETS. In contrast, only LNG used in diesel slow-speed engines will benefit from excess gains under both regulatory frameworks.

In 2030, while fossil-based LNG, Otto engines and diesel slow-speed options, along with LPG alternatives (Butane and Propane), will continue to generate a surplus under FuelEU, they will face penalties under the IMO framework. They will be subject to exposure to Tier 1 RUs. The same applies to the assumed Bio24 (24% Biodiesel plus VLSFO) and Bio30 (30% Biodiesel) options considered in the study. Only the Bio100 fuels, which achieve 65% and 80% emission reductions, will generate a surplus under both regulations.

Bio100 is the only compliance option available for 2035. In contrast, the IMO’s regulation allows banking only for two years, while FuelEU permits indefinite banking. This indicates that the IMO’s framework is notably more stringent regarding LNG-powered vessels, necessitating an earlier transition to bio-LNG or e-LNG.

Table 2. FuelEU and GFS regimes.

Aspect	FuelEU	GFS
Enforcement structures	Direct, legally binding requirements under EU law, which are enforced by member states and monitored by accredited third-party verifiers.	The framework depends on flag-state enforcement through the MARPOL Annex VI regulations. The compliance will be confirmed by the IMO-maintained Global Fuel Intensity (GFI) registry.
Economic incentives and penalties	FuelEU functions solely as a penalty-driven model.	GFS closely resembles FuelEU in structure, but it incorporates a global pricing system.
Pricing models	FuelEU imposes a fixed penalty of EUR 2400 for each metric ton of VLSFO-equivalent shortfall, irrespective of greenhouse gas intensity performance or global market conditions.	Under the IMO framework, Tier 1 remedial units are valued at USD 100 per metric ton of CO2eq, while Tier 2 remedial units are priced at USD 380 per metric ton.
Political drivers	FuelEU was created as part of the EU’s Fit for 55 legislative initiatives. It allows for only limited options for withdrawal or modification, requiring new actions from both the Commission and the Parliament.	IMO moved away from the consensus decision-making process and, by voting during MEPC (83), reached an agreement in response to global trade tensions.
Jurisdiction	Regional–Applies to ships using EU ports.	Global–Applies to all IMO member states.
Applicability	This applies to ships over 5000 GT operating within the EU, excluding non-commercial vessels, government vessels, and those that use non-mechanical propulsion systems.	Global application of MARPOL for vessels exceeding 5000 GT, with exceptions that may apply based on flag state, geographic region, and type of propulsion.
Target Year for Implementation	The regulations will take effect in 2027, with the first reporting period beginning in 2028.	This takes effect from 1 January 2025.
Fuel Certification	Fuel suppliers are required to adhere to EU fuel certification regulations, including RED II and FuelEU templates.	The GHG profile of the fuel must be certified according to the IMO Lifecycle Guidelines.

presents a comparison of both GHG reduction regimes across various aspects.

While technical standards may gradually align, the underlying structural and political differences between the two regimes are likely to remain. As a result, shipowners will need to implement dual compliance strategies that cater to their specific operating patterns and market conditions.

3. GHG Intensity for the Main Ship Types According to FuelEU

The information utilized for the analyses is sourced from the THETIS-MRV database (https://mrv.emsa.europa.eu/#public/emission-report accessed on 25 July 2025). Each vessel’s data encompasses the following:

• Total fuel consumption [m ton]–FC.
• Total ${CO}_2$ emissions [m ton]–${CO}_2$.
• ${CO}_2$ emissions which occurred within ports under an MS jurisdiction at berth [m ton]–${CO}_2^{port}$.

The study used available data from 2020 to 2023. The first version of the emissions data in the EU-MRV database for 2024 was published on 30 June 2025. The number of vessels for every year and type from the database is shown in

Table 3. Ships from database.

Ship Type	2020	2021	2022	2023
General cargo ship	1226	1237	1220	1175
Bulk carrier	3566	3553	4067	3624
Tanker	1920	1994	2006	1908
Container ship	1827	1716	1854	1916

.

The total fuel consumption (FC) is divided into two types, VLSFO and MDO, under the following assumptions (14)–(16):

$$M_{VLSFO}+M_{MDO}=FC$$

(14)

$$M_{VLSFO}\times {EF}_{VLSFO}+M_{MDO}\times {EF}_{MDO}={CO}_2$$

(15)

where

$$M_{MDO}\ge {CO}_2^{port}/{EF}_{MDO}$$

(16)

where $M_{HFO}$ is the mass of VLSFO (m ton), $M_{MDO}$ is the mass of MDO (m ton), ${EF}_{VLSFO}$ is the emission factor for VLSFO (3.114), and ${EF}_{MDO}$ is the emission factor for MDO (3.206).

Additionally, information on the deadweight and age of each ship by the end of 2024 was obtained from the Marine Traffic website (https://www.marinetraffic.com/ accessed on 3 August 2025). The youngest is the fleet of tankers included in the database.

Table 4. Average age and DW of ships from EU-MRV for 2023.

Ship Type	Age (Avg), Year	DW (Avg), t
General cargo ship	15.8	21,779.8
Bulk carrier	11.4	65,058.1
Tanker	10.8	111,305.2
Containership	14.8	80,260.6

presents the average age and DW of the ships from EU-MRV for 2023. The youngest is the fleet of tankers included in the database.

Using the data on fuel consumption for each year, GHG intensities and penalties for non-compliance for 2025 were calculated according to FuelEU (1)–(7). The total amounts of CO₂eq and the corresponding penalties are presented in Figure 2.

The ${GHG}_{intensity}$ of VLSFO and MDO(MGO) is 91.60 $\mathrm{g}{\mathrm{CO}}_{2\mathrm{eq}}/\mathrm{MJ}$ and 90.63 $\mathrm{g}{\mathrm{CO}}_{2\mathrm{eq}}/\mathrm{MJ},$ respectively. Therefore, for all ships in the data for each of the years, the GHG intensity will be between these two values, depending on the percentage ratio of the two fuels consumed. None of the ships can meet the 2025 requirement of 89.3368 $\mathrm{g}{\mathrm{CO}}_{2\mathrm{eq}}/\mathrm{MJ}.$

The relation between the average GHG intensity based on 2023 data and the age of ships is presented in Figure 3. The lower average values of GHG intensity for older general cargo ships, bulk carriers, and tankers are associated with ships that have a smaller deadweight and primarily use diesel fuel.

4. Contour Line Analysis

Haver’s contour technique [27] is used to identify extreme responses. It is applied in this study to assess GHG emissions estimated based on the methodology outlined in Annex I of the European regulation [9] from four types of ships (tankers, bulk carriers, general cargo ships, and container ships) visiting European ports, relying on data from the EU-MRV. First, the method establishes a model that describes the combined distribution of operational control factors, assumed here as DW and ship age. This joint distribution is then utilized to identify a set of operational characteristics that share the same joint exceedance probability, which forms a contour.

Next, GHG emissions generated by the analyzed ships are calculated for specific operational characteristics along the contour, referred to as limited operational conditions.

The maximum GHG emissions found along this contour are compared to the maximum allowable emissions stipulated by the EU as ${GHG}_{limit,EU2020}$ = 91.16 gCO_2eq/MJ.

The contour approach [28] identifies a “Contour” within the region covered by the operational parameters x₁ and x₂, where GHG emissions meet all criteria, ensuring that the probability of crossing does not exceed an “Unacceptable limit”. The crossing point of the “Contour” and the “Unacceptable limit” defines the mean extreme value of the operational parameters as can be seen in Figure 4.

Different methods for defining line contours are based on various assumptions regarding the shape of unacceptable limits—specifically, the region in the control parameter space where GHG emissions are deemed unacceptable. As a result, the definition of what constitutes an exceedance of the contour also varies.

In this context, based on the collected data, it is assumed that ship age and deadweight tonnage (DW) are the primary operational control parameters for the present study. GHG emissions are classified as unacceptable whenever the operational conditions fail within this unacceptable region.

The Inverse First-Order Reliability Method, IFORM [28,29,30], commonly used in various fields, assumes that the unacceptable region is convex. This assumption means that the probability of contour exceedance is considered as a marginal exceedance probability after adjusting for the axis rotation.

The GHG emissions, based on the data collected, depend on control operational variables—specifically, ship age and deadweight tonnage. Emissions are regarded as unacceptable whenever the operating conditions fail within this GHG unacceptable region. The contour exceedance probability can be defined as a marginal exceedance probability when the axes are rotated.

The First-Order Reliability Method, FORM [31,32], which can provide a conservative estimate of the failure probability when the unacceptable area is convex, is employed. The contour exceedance probability has been adjusted to account for short-term variations in greenhouse gas emissions.

The likelihood of greenhouse gas emissions being outside a contour is greater than the marginal exceedance probability, which refers to the chance that the emissions generated by a ship exceed the contour’s upper limit in each dimension.

Employing the IFORM contour for computing overall exceedance probabilities permits identifying the limits for GHG emissiozns, acting as an upper bound for the possibility of crossing that contour. The total exceedance probability of a contour is utilized to evaluate the accuracy of a joint probability model in relation to the data, as it provides an estimate for the expected number of occurrences outside the contour.

The FORM is used to estimate the probability of crossing the GHG unacceptable limit (see Figure 5). The joint distribution is transformed into independent standard normal variables through the Rosenblatt transformation [33]. The design point is identified as the point with the highest probability on the failure surface in U-space. This point corresponds to the location on the GHG unacceptable surface that is closest to the origin.

The reliability index, denoted as ${\beta }_F$, represents the distance from the origin to the design point. It calculates the probability that greenhouse gas emissions will exceed the unacceptable GHG limit, and it is linear at the design point.

The IFORM contour in U-space is defined as the set of points that are at a distance of ${\beta }_F$ from the origin. By utilizing the inverse Rosenblatt transformation, this contour in U-space can be mapped back to the original parameter space.

In U-space, the marginal exceedance probability corresponds to the maximum values of the standard normal variables within the IFORM contour, given a specified exceedance probability.

The Weibull distribution [34] is utilized to ascertain the marginal distribution of the deadweight:

$$F_{DW}=1-\mathrm{e}\mathrm{x}\mathrm{p}\left[-{\left({\displaystyle \frac{DW-{\gamma }_{DW}}{{\alpha }_{DW}}}\right)}^{{\beta }_{DW}}\right]$$

(17)

where ${\alpha }_{DW}$ is the scale, ${\beta }_{DW}$ is the shape, and ${\gamma }_{DW}$ is the location parameter.

The ship’s age, $A_S$, is modeled conditionally on $DW$ by a Lognormal distribution:

$$f_{A_S|DW}={\displaystyle \frac{1}{{\sigma }_{A_S}{A}_S\sqrt{2\pi }}}exp\left[-{\displaystyle \frac{\left(ln{A}_S-{\mu }_{A_S}\right)}{2{\sigma }_{A_S}^2}}\right]$$

(18)

where the distribution parameters ${\mu }_{A_S}$ and ${\sigma }_{A_S}$ are functions of the deadweight. A good fit to the data can be achieved as follows:

$${\mu }_{A_S}\left(DW\right)=a_0+a_1{DW}^{a_2}$$

(19)

$${\sigma }_{A_S}\left(DW\right)=b_0+b_1{e}^{b_{2}DW}$$

(20)

where the coefficients $a_i,b_i$ are defined using data from ships visiting European ports.

The distribution of the annual maximum [35] deadweight, $F_{{DW}_{max,1yr}}$, is defined as follows:

$$F_{{DW}_{max,\mathrm{ }\mathrm{ }1yr}}\left(DW\right)={F_{DW}\left(DW\right)}^n$$

(21)

where $n$ is the number of ships that visited the European ports in one year. The $DW$ with a return period, $T_R$, is defined as follows:

$${DW}_{T_R}={F_{DW}}^{-1}\left(1-{\displaystyle \frac{1}{n{T}_R}}\right)$$

(22)

The contour concept is utilized to establish the extreme GHG emission condition, in which the joint model for $DW,A_S$ is defined as follows:

$$f_{DW,A_S}\left(DW,A_S\right)=f_{DW}\left(DW\right)f_{A_S|DW}\left(A_S\right)$$

(23)

and the conversion to the standardized normalized U-space results in

$$\mathsf{\Phi }\left(u_1\right)=F_{DW}\left(DW\right),\mathsf{\Phi }\left(u_2\right)=F_{A_S|DW}\left(A_S\right)$$

(24)

and

$$\sqrt{u_1^2+u_2^2}={\beta }_F$$

(25)

where ${\beta }_F$ is the radius for the $n$-year contour.

The transformation of the circle into a contour in the parameter space is achieved through the following:

$$DW={F_{DW}\left(\mathsf{\Phi }\left(u_1\right)\right)}^{-1}{A}_S={F_{A_S|DW}\left(\mathsf{\Phi }\left(u_2\right)\right)}^{-1}$$

(26)

The probability of a point with GHG emissions falling within a contour of radius ${\beta }_F$, is expressed as follows:

$$P_{r,\mathrm{ }\mathrm{ }in}=\iint \mathsf{\Phi }\left(u_1,u_2\right)d{u}_{1}d{u}_1=1-e^{-{\displaystyle \frac{1}{2}}{{\beta }_F}^2}$$

(27)

and the anticipated quantity of points within the contour is defined as follows:

$$n_{in}=n_{a}Y\left(1-e^{-{\displaystyle \frac{1}{2}}{{\beta }_F}^2}\right)$$

(28)

where $n_a$ is the annual average number of ships visiting European ports, and $Y=1$ yr is the year the data of the present study covers.

The contour represents extreme operational conditions, depicted as a curve that indicates equal probability for specific ${\beta }_F$ levels. These levels are determined by a joint function of the ship’s deadweight and age.

The contour in operational parameter space can be seen for the four types of ships visiting European ports in 2023 analyzed when ${\beta }_F$ = 1, 2, and 3 in Figure 6, showing a very good fit of the observed data to the contours, including the probability of a point with GHG emissions being inside the contour, the exceedance probability, the expected number of points inside the contours, and the reliability index, representing the distance from the origin to the design point.

These contours present the levels of greenhouse gas emissions, conditional on the statistical descriptors of the contours. If a specific deadweight characterizes an operational condition and the ship age is located within this contour, it is deemed acceptable. Conversely, if the operational condition lies outside the contour, it represents a condition that leads to unacceptable environmental impacts.

The ${\beta }_F$ levels define these extreme operational conditions independently of the ship’s design. Ships respond differently under extreme conditions, which can vary along the contour. There is a point on the contour curve where GHG emissions are at their peak, resulting in the most significant environmental impact. The maximum emissions depend on the operational conditions, which are independent yet represented on the same contour.

When assessing the energy efficiency of a ship, it is crucial to ensure that greenhouse gas emissions remain below the GHG unacceptable limit. In this regard, the unacceptable limit is defined based on GHG emissions according to FuelEU and the corresponding levels of ${GHG}_{limit,EU2020}$ = 91.16 gCO_2eq/MJ for ships visiting European ports in 2023. The unacceptable limit was derived based on collected data for the operational control parameters $DW$ and $A_S$ for any type of ship as a power function:

$$DW\left(As\right|{GHG}_{limit})=a\times {\left(A_S\right)}^b$$

(29)

where $a$ and $b$ are established for all analysed ships. After defining the $DW\left(As|{GHG}_{limit}\right)$, the contour is adjusted to meet the limit, resulting in the identification of extreme values characteristic of the operational parameters and the acceptable region, as can be seen in Figure 7.

As the deadweight increases, reaching the extreme mean value, ship resistance also rises, thus requiring more propulsion power. At the same time, the age of the vessel significantly affects energy efficiency. Any combination of deadweight and ship age can result in various levels of GHG emissions.

The analysis conducted in this study indicates that the operational conditions associated with the contour line will identify GHG emissions that are within the limits set by the ${GHG}_{limit,EU2020}$ = 91.16 gCO_2eq/MJ for ships visiting European ports in 2023. The descriptors of the contour lines for diverse types of ships are presented in

Table 5. Contour ${GHG}_{limit,EU2020}$ = 91.16 $\mathrm{g}{\mathrm{C}\mathrm{O}}_2\mathrm{e}\mathrm{q}/\mathrm{M}\mathrm{J}$ descriptors.

Ship Type	${\mathit{n}}_{\mathit{i}\mathit{n}}$	Q	${\mathit{\beta }}_{\mathit{F}}$	${\mathit{D}\mathit{W}}_{\mathit{e}\mathit{x}\mathit{t}\mathit{r}}$, Metric Ton	${\mathit{A}}_{\mathit{S},\mathit{e}\mathit{x}\mathit{t}\mathit{r}}$, Years
Bulk carrier	1041 (29%)	0.202	0.8	56,557	5.56
Container ship	352 (18%)	0.257	0.7	104,321	5.89
Tanker	445 (23%)	0.229	0.7	149,519	3.25
General cargo ship	568 (48%)	0.123	1.2	25,430	10.21

.

It can be identified that the most energy-efficient ships concerning the ${GHG}_{limit,EU2020}$ are the general cargo vessels, with a probability of exceeding greenhouse gas emissions within the contour of Q = 0.123. The expected number of ships visiting European ports that meet the allowable emissions limit is $n_{in}$ = 568, which is 48% of the total number of general cargo ships that visited European ports in 2023, and the reliability index, which represents the distance from the origin to the limit, is ${\beta }_F$ = 1.2. The extreme mean values are $A_{S,extr}$ = 10.21 years and ${DW}_{extr}$ = 25,430. The least energy-efficient type of ship is the bulk carrier.

Utilizing contour line formulation to analyze the GHG emissions generated by ships visiting European ports in 2023 highlights the significance of this study in evaluating the mean values and efficiency of any ship. Although this study is currently limited to a specific set of operational parameters, it can be expanded to include additional relevant factors.

5. Conclusions

This study examined the impact of the FuelEU Maritime Regulation on four types of vessels that frequently visit European ports over four years. It utilized relevant emission data obtained from the EU MRV database. For all ships using only VLSFO and MDO, the GHG intensity calculated from data over the past four years (2020–2023) falls between the WtW GHG emission factors of both fuels, which range from 91.60 to 90.63 gCO_2eq/MJ. This variation depends on the ratio of each fuel consumed. None of the ships can comply with the 2025 requirement of 89.3368 gCO_2eq/MJ.

Additionally, the study discussed the recent amendments to MARPOL Annex VI, which include the establishment of a Global Fuel Standard (GFS) by the International Maritime Organization (IMO) to lower greenhouse gas emissions. A comparison was made between the GFS and the existing FuelEU Regulation. The GHG emissions regulations set by the EU and the IMO aim for the same objective, but they differ in the speed of GHG reduction and include various additional conditions. There is a risk of simultaneously applying the two groups of rules, and it is expected that after October of this year, there will be clarity and harmonization. When analyzing the use of alternative fuels, it is essential to consider the requirements of both the EU and the IMO. This includes not only assessing the increased costs of alternative fuels but also evaluating the benefits of reduced penalties and potential surplus opportunities. A viable alternative is the use of biodiesel blended fuels (Bio30), which should be analyzed after the final adoption of the IMO requirements and based on 2024 data available in the EU-MRV database as of 30 June 2025.

The contour line approach was utilized to analyze the energy efficiency of ships visiting European ports, relying on data from the EU-MRV. The developed generic model can be applied to any vessel. The greenhouse gas emissions for the short-term response was estimated using the methodology for the estimation of GHG intensity, outlined in Annex I of [9]. The use of contour lines to analyze GHG emissions underscores the significance of this study in assessing a ship’s energy efficiency. Specifically, the analysis that considered ${GHG}_{limit,EU2020}$ = 91.16 gCO_2eq/MJ revealed that general cargo ships are the most energy-efficient type, while bulk carriers are the least efficient. This approach allows for analyzing the operational profile of the ship and her energy efficiency. The study was conducted using real data, and the contour approach was instrumental in evaluating GHG emissions, marking a significant advancement in the methodology, modelling, and analysis of a ship’s energy efficiency.