What is EEDI? Energy Efficiency Design Index

What EEDI is and why it exists

The Energy Efficiency Design Index is a single number that tells you how many grams of carbon dioxide a new ship emits to move one tonne of cargo one nautical mile. The IMO adopted it through Resolution MEPC.203(62) on 15 July 2011, the first mandatory carbon-intensity standard ever applied to an entire transport sector. It binds at the drawing board, not at sea. A naval architect computes the attained EEDI from the propulsion plant, the fuel, the hull, and the design speed, then proves that number sits at or below the required EEDI for that ship type, deadweight, and contract date.

Before 2011 there was no design-phase carbon metric for ships. An owner could order any hull form, any engine, any propeller, and class would certify it for strength and stability with no view on its CO2 per tonne-mile. International shipping burns roughly 250 to 300 million tonnes of fuel oil a year and accounts for close to 3 percent of global anthropogenic CO2, a share the IMO’s Fourth GHG Study (2020) put at 2.89 percent for 2018. EEDI was the lever the IMO chose to bend the newbuild end of that curve, on the logic that a ship ordered today will still be trading in 2050.

The legal home of EEDI is Chapter 4 of MARPOL Annex VI, the air-pollution annex. MEPC.203(62) inserted that chapter, and the amendments entered force on 1 January 2013 under MARPOL’s tacit-acceptance procedure. Regulation 19 sets the application scope, Regulation 20 defines the attained EEDI, Regulation 21 defines the required EEDI, and Regulation 24 carries the table of ship types and their reduction phases. The 2021 consolidation in MEPC.328(76) renumbered and tightened these provisions while folding in the EEXI requirement for existing ships, so a current copy of Annex VI reads slightly differently from the 2011 text even though the EEDI machinery is the same.

How EEDI got onto the statute book

EEDI didn’t arrive fully formed. The IMO’s Marine Environment Protection Committee worked the design-index idea through most of the 2000s, starting from a voluntary trial metric called the EEDI interim guidelines circulated as MEPC.1/Circ.681 in 2009, alongside the operational EEOI guidance in MEPC.1/Circ.684. Owners and yards ran the trial formula on real ships and fed the results back, which is how the IMO got the 1999-to-2009 data set it later regressed into the reference lines. The trial period mattered: it turned a paper formula into one calibrated against the actual fleet.

The committee adopted the mandatory measure at MEPC 62 in July 2011 through Resolution MEPC.203(62), inserting Chapter 4 into MARPOL Annex VI. That adoption was a fight. A bloc of developing-country delegations argued the measure cut against the UN climate principle of common-but-differentiated responsibilities, and the resolution passed by vote rather than consensus, an unusual step at the IMO, with several states recording reservations. The amendments entered force on 1 January 2013 under the tacit-acceptance rule, which deems a MARPOL amendment accepted unless enough flag states object by a set date.

The detail then came in supporting resolutions over the next two years. MEPC.212(63) and MEPC.213(63) at MEPC 63 in March 2012 set out the calculation method and the SEEMP requirement; MEPC.214(63) at the same session set the survey-and-certification guidelines. MEPC 65 in May 2013 replaced the calculation guidance with MEPC.231(65), the version still cited today as amended, and added the minimum-power guidelines in MEPC.232(65). Later sessions widened the net: MEPC.251(66) in 2014 brought in LNG carriers and ro-ro and cruise ships with non-conventional propulsion, and MEPC.281(70) in October 2016 revised the calculation method and reference lines for ro-ro and ro-pax types. The 2021 consolidation in MEPC.328(76) gathered Chapter 4 and the new EEXI provisions into a single revised Annex VI that took effect on 1 November 2022.

The attained EEDI: CO2 out over transport work done

The attained EEDI answers one question. Run the ship at its reference condition, count the CO2 its engines produce in an hour, divide by the useful transport work it does in that hour, and what’s the ratio? The reference condition is fixed so two ships can be compared: 75 percent of the maximum continuous rating of the main engines, on the summer load draft, in calm water. That 75 percent MCR point is meant to stand in for a realistic service load rather than the engine’s flat-out rating.

Read the numerator and denominator in turn. The numerator is mass of CO2 per hour. For the main engines it’s the power $P_{ME}$ (75 percent of total installed MCR after deducting any shaft-generator and innovative-technology power) times the specific fuel consumption $\text{SFC}_{ME}$ from the engine’s NOx Technical File, times the fuel carbon factor $C_{f}$. The auxiliary load $P_{AE}$ is estimated from a formula in MEPC.231(65) keyed to total propulsion power, and it carries its own SFC and Cf. So the numerator is just kilowatts times grams-per-kilowatt-hour times tonnes-CO2-per-tonne-fuel, summed across the prime movers.

The carbon factor $C_{f}$ is not a guess. MEPC.364(79) tabulates it by fuel: 3.114 t CO2 per t fuel for diesel and gas oil, 3.206 for heavy fuel oil, and 2.750 for LNG, among others. The denominator is transport work: capacity times reference speed $V_{ref}$. Capacity is deadweight for most cargo ships and gross tonnage for passenger and ro-pax types, a distinction set ship-by-ship in the Regulation 24 table. $V_{ref}$ is the speed the ship makes at the 75 percent MCR condition, taken from tank tests and confirmed at the sea trial.

Power, shaft generators, and the PTI/PTO terms

The full MEPC.231(65) numerator is more involved than the single-engine sketch, because real ships move electrical power around between the main engine and the auxiliary plant. The method handles four power paths. $P_{ME}$ is 75 percent of total main-engine MCR, less the power drawn by any shaft generator and any innovative-technology contribution. $P_{AE}$ is the auxiliary load, either estimated from the deadweight-keyed formula when total propulsion power exceeds 10,000 kW, or measured for smaller plants. A power take-off, where a shaft generator draws off the main engine to make electricity, reduces the main-engine power available for propulsion and so enters the formula through $P_{PTO}$. A power take-in, where a shaft motor adds propulsion power from auxiliary generators, enters through $P_{PTI}$ scaled to the propulsion side. Getting these signs right is where most first-pass EEDI calculations go wrong, because a shaft generator that looks like a fuel saving in one term can add auxiliary fuel in another.

The fuel side carries its own complications. A dual-fuel engine running on LNG with a pilot diesel injection has two carbon factors and two SFC values in the same engine term, weighted by the pilot-fuel fraction. A ship with separate fuels for main and auxiliary engines, common with an HFO main and an MGO auxiliary, carries different $C_f$ values in each. The specific fuel consumption figures aren’t nominal catalogue numbers either: they come from the engine’s NOx Technical File, certified under the NOx Technical Code, and corrected to the reference load and to a lower-calorific-value basis of 42,700 kJ/kg so that engines on different fuels compare fairly.

Two ships of the same size and speed can land at very different attained EEDI values, and the formula doesn’t care how. A bulk carrier moving 82,000 tonnes at 14.2 knots might hit its target through a fuller hull and a larger-diameter slow-turning propeller; another might use a derated engine and a shaft generator; a third might add a wind-assist rotor system. EEDI is technology-neutral by design: it scores the outcome in g CO2 per tonne-mile, and the designer picks the path.

Correction factors that bend the headline number

The bare ratio above is the single-engine, single-fuel, no-shaft-generator case. The full MEPC.231(65) method wraps it in a set of correction factors that account for design features the basic ratio would penalize unfairly. These factors multiply capacity, speed, or the power terms, and getting them right is most of the real work in an EEDI calculation.

The capacity-side factors adjust the denominator. $f_i$ is the capacity correction for technical or regulatory limits on deadweight, used mainly by ships with ice-strengthened hulls that carry heavier structure and so lose cargo capacity. $f_c$ is the cubic-capacity correction for chemical tankers and gas carriers whose volume-limited holds mean their deadweight understates their real carrying job; the cubic-capacity correction lifts the effective capacity when the ratio of cubic capacity to deadweight crosses a threshold. The $f_{vse}$ factor credits ships that carry voluntary structural enhancement above class minimums, again because that steel costs deadweight.

The power-side factors adjust the numerator or the speed. $f_j$ is the ship-specific power correction. For ice-classed ships it discounts the extra installed power needed to break ice, since penalizing a Baltic trader for the engine it needs in February would be perverse; the ice-class correction reads off the ice class (IA Super down to IC) and the ship type. $f_w$ is the weather factor, a voluntary coefficient below 1.0 that credits a ship for the speed it can still hold in representative wind and wave conditions, measured against the MEPC.1/Circ.796 standard. And $f_m$ corrects for ice-classed ships with diesel-electric, turbine, or hybrid propulsion. Each factor traces to a specific paragraph of MEPC.231(65), and a surveyor verifying the EEDI Technical File checks the derivation of every one.

The required EEDI: a declining reference line

The attained value is only half the test. The other half is the required EEDI, the ceiling the ship must beat. The IMO built that ceiling from real fleet data. For each ship type it regressed the EEDI of ships built between 1999 and 2009 against deadweight, producing a reference line, then set the requirement as a fixed percentage cut below that historical baseline.

The reference line is a power curve, $a \times \text{Capacity}^{-c}$, where $a$ and $c$ are type-specific constants tabulated in MEPC.328(76). Larger ships sit lower on the curve because they’re inherently more efficient per tonne-mile, so the required value for a 200,000 DWT bulk carrier is far smaller than for a 25,000 DWT one. The negative exponent $c$ captures that economy of scale. You can step through the curve coefficients for any type and size with the reference-line calculator; the full attained-versus-required check, including the compliance margin, runs in the required-EEDI calculator.

The reference-line constants were set in MEPC.231(65) and carried forward, with adjustments, into the consolidated Annex VI. For bulk carriers the line uses $a = 961.79$ and $c = 0.477$; for tankers $a = 1218.80$ and $c = 0.488$; for container ships $a = 174.22$ and $c = 0.201$. Gas carriers, LNG carriers, general cargo, refrigerated cargo, combination carriers, ro-ro cargo, ro-ro passenger, and cruise passenger each carry their own pair. Two values, one regression, and the historical average efficiency of an entire ship class falls out as a single curve.

Phase reduction factors: 0, 10, 20, 30 and steeper

The reference line is the year-2000 baseline. The required EEDI tightens below it in steps through the reduction factor $Z$, the percentage cut applied to the line. The phases ratchet by contract date, so the longer you wait to order a ship, the harder the target.

Phase 0 ran from 1 January 2013 to 31 December 2014 with a 0 percent cut: ships had to match the reference line, no better. Phase 1 ran 2015 to 2019 at a 10 percent cut. Phase 2 ran 2020 to 2024 at 20 percent for most types. Phase 3 took effect for ships contracted from 1 January 2025 at 30 percent for most types, but the IMO pulled that date forward for the heaviest emitters. MEPC.324(75), adopted November 2020 and folded into the consolidated Annex VI, moved container ships, gas carriers, general cargo ships, and LNG carriers into Phase 3 from 1 April 2022, and for the largest container ships, those above 200,000 DWT, set the Phase 3 cut as high as 50 percent rather than 30. The size-band detail matters: a 15,000-TEU box ship faces a steeper required reduction than a 1,000-TEU feeder of the same vintage. The phase-factor calculator resolves $Z$ for a given type, size band, and contract date.

What does 30 percent mean on the water? A 2025-built tanker has to emit roughly 30 percent less CO2 per tonne-mile than a tanker matching the year-2000 fleet average. That gap gets closed through some mix of hull optimization, larger and slower propellers, derated or right-sized engines, waste-heat recovery, and in a growing share of orders, dual-fuel engines burning LNG, methanol, or ammonia. EEDI is the quiet reason newbuild design speeds have crept down over the past decade: a slower design speed lowers $V_{ref}$ in the denominator but lowers the cubed power demand in the numerator far faster, so the ratio falls.

A worked sense of the numbers

Put rough figures through the ratio to see how it behaves. Take a Handymax bulk carrier with 9,000 kW total main-engine MCR, so $P_{ME}$ is 6,750 kW at the 75 percent point, burning heavy fuel oil at an SFC of 175 g/kWh. The HFO carbon factor is 3.114 t CO2 per t fuel under MEPC.364(79). The main-engine CO2 rate is 6,750 times 175 times 3.114, about 3.68 million grams of CO2 per hour. Add a 600 kW auxiliary load at 215 g/kWh on marine gas oil at a 3.206 factor, roughly another 0.41 million grams per hour, for a numerator near 4.09 million g CO2 per hour.

Now the denominator. Say the ship carries 56,000 tonnes deadweight at a reference speed of 14.5 knots. Transport work is 56,000 times 14.5, about 812,000 tonne-nautical-miles per hour. Divide: 4.09 million over 812,000 lands the attained EEDI near 5.0 g CO2 per tonne-mile. The bulk-carrier reference line, $961.79 \times 56000^{-0.477}$, sits around 5.5, so before any phase cut this ship is comfortably under the year-2000 baseline. Apply the Phase 3 30 percent reduction and the required value drops to about 3.9, at which point the same ship is over the limit and the designer has to find the gap through a slower design speed, a larger propeller, waste-heat recovery, or a wind-assist credit. The numbers here are illustrative, not certified; the attained-EEDI calculator and the required-EEDI calculator run the full method with the correct factors.

Ship types in scope, and what’s excluded

EEDI applies to ships of 400 gross tonnage and above engaged on international voyages, in the ship types listed in Regulation 2 and tabulated in Regulation 24 of the consolidated Annex VI. The covered types are bulk carriers, gas carriers, tankers, container ships, general cargo ships, refrigerated cargo carriers, combination carriers, ro-ro cargo ships, ro-ro cargo ships carrying vehicles, ro-ro passenger ships, LNG carriers, and cruise passenger ships with non-conventional propulsion. Each type has its own reference-line constants and, for some, its own phase schedule.

The application threshold has a wrinkle worth knowing. The EEDI calculation requirement bites at 400 GT, but the survey-and-certification regime that produces the certificate has been extended over time, and several ship types were added to the scope only after the original 2013 entry. Ro-ro and ro-pax ships, for instance, came under EEDI through MEPC.281(70) in October 2016, with reference lines and phase dates set later than for the original cargo types. LNG carriers and cruise ships with diesel-electric or hybrid propulsion were brought in through MEPC.251(66) in 2014. The scope grew because the original 2011 reference lines only covered the ship types for which the IMO had enough 1999-to-2009 data to regress a reliable curve.

Exclusions are explicit. Ships with diesel-electric, turbine, or hybrid propulsion in types other than those listed are outside the index, as are ships not propelled by mechanical means, platforms including floating production storage and offloading units and floating storage units, drilling rigs, and ships of categories not assigned a reference line. Warships, naval auxiliaries, and government non-commercial ships are excluded by Annex VI Regulation 3, the standard MARPOL sovereign-vessel carve-out. Ships engaged solely on domestic voyages fall to flag-state discretion rather than the international requirement.

Minimum propulsion power: the safety floor under EEDI

EEDI rewards low installed power, because lower power means lower numerator. Taken to an extreme, that incentive could push a designer to under-power a ship to the point where it can’t hold heading or steerage in a storm, which is a different way to lose the ship and its crew. The IMO saw this coming and built a floor.

The minimum propulsion power requirement says a ship’s installed power can’t fall below what it needs to maintain maneuverability in adverse weather. The original framework came in MEPC.232(65) as 2013 interim guidelines, refined through MEPC.255(67) and later revisions, and traces back through circular MEPC.1/Circ.850 and its successors. The guidelines give two assessment levels: a minimum-power-line check, where installed power must exceed a deadweight-keyed line value, and a more detailed simulation-based assessment for ships that fail the line but can demonstrate maneuverability in defined wind and wave conditions. Annex VI Regulation 21 makes meeting one of these levels a condition of EEDI compliance for tankers, bulk carriers, and combination carriers, the types most exposed to under-powering.

This is the part of EEDI most argued over. Class societies, owners, and the IMO’s own correspondence groups have debated for years whether the minimum-power lines are set tightly enough, because a ship that just clears the line on paper can still struggle in the real North Atlantic. The tension is structural: EEDI pushes power down, the safety floor pushes it back up, and the floor’s exact height is a regulatory judgment that’s been revised more than once and remains a live agenda item at MEPC sessions.

Innovative energy-efficient technologies

EEDI credits technologies that cut CO2 in ways the basic engine-power calculation would miss. MEPC.231(65) and circular MEPC.1/Circ.815 sort these into three categories, and each enters the formula at a different point.

Category A is technologies that reduce the ship’s own energy demand without generating power, such as low-friction hull coatings or hull-air-lubrication systems that cut frictional resistance. These shift the effective power through the $(1 - f_{eff})$ term. Category B is technologies that generate auxiliary electrical power that would otherwise come from burning fuel, mainly waste-heat-recovery systems and shaft generators driven off the main engine; their output is subtracted from the auxiliary load in the numerator. Category C is technologies that generate propulsion power directly, the headline case being wind propulsion: rotor sails, rigid wing sails, kites, and conventional sails. Their contribution is subtracted from the main-engine power, weighted by an availability factor $f_{eff}$ that reflects how often the wind actually does the work over a representative route.

The availability factor is the technical heart of the wind credit, and MEPC.1/Circ.815 sets out how to compute it. For a wind-propulsion system the credit depends on a global wind matrix, the wind speed and direction probabilities along a representative route, combined with the system’s thrust characteristics at each wind angle. A Flettner rotor produces most of its thrust on a beam reach and little dead downwind or upwind, so its $f_{eff}$ reflects how often the modelled route puts wind on the beam. The circular requires the system supplier to document the thrust curve and the route assumptions, and the verifier checks them, because an optimistic wind matrix would inflate the credit just as an optimistic design speed would inflate the headline ratio.

The category split also decides where uncertainty falls. Category B systems like waste-heat recovery are relatively easy to verify because their output is measured electrical power on a known duty cycle, so the credit is firm. Category C wind systems carry more modelling uncertainty because the resource varies with route and season, which is why the availability factor and the route documentation matter so much. The IMO has revised the innovative-technology guidance more than once as installations multiplied, and the treatment of wind in particular has been a recurring item, since the early circulars predate the current generation of large rotor and wing-sail retrofits.

The wind credit is where EEDI is reshaping orders. A rotor-sail installation that delivers a documented average power saving along a defined trade lane lets a designer either lower installed engine power or accept a higher design speed while still clearing the required EEDI. The availability factor keeps the credit honest: a kite that only pulls on a downwind leg can’t claim full-time propulsion. The innovative-technology calculator shows how a given credit moves the attained value against the required ceiling.

Verification: from preliminary EEDI to the IEE Certificate

EEDI is verified twice, once on paper and once on the water. The process and the documents are set out in MEPC.214(63), the 2012 survey-and-certification guidelines, read with Annex VI Regulation 5 on surveys.

The first step is the preliminary verification at the design stage. The shipbuilder submits an EEDI Technical File to the verifier, usually the classification society acting as a recognized organization for the flag state. The Technical File carries the full attained-EEDI calculation: the power figures, the SFC values from the NOx Technical File, the carbon factors, the capacity, the estimated reference speed from tank tests, every correction factor with its derivation, and any innovative-technology credit. The verifier checks the method against MEPC.231(65) and issues a preliminary attained EEDI. This is the number the ship is designed to.

The sea-trial correction is more involved than it sounds, and it’s where verifications most often slip. A speed trial run in any real sea state has to be corrected back to the ideal reference condition: no wind, no waves, deep water, the contractual displacement. MEPC.231(65) points to the ISO 15016 method for these corrections, which back out the added resistance from waves, the wind drag on the hull above the waterline, the effect of water temperature and density on resistance, and the shallow-water effect when the trial site isn’t deep enough. Each correction shifts the measured speed-power curve, and the corrected curve, not the raw trial run, sets the verified $V_{ref}$. A verifier who accepts a poorly corrected trial can certify a speed the ship can’t actually make in the reference condition, which is one route to an EEDI that looks better on the certificate than in service.

The second step is the final verification at the sea trial. The ship runs its speed trials, the measured speed-power curve is corrected to the standard reference condition for sea state, wind, displacement, and shallow water using the methods in MEPC.231(65), and the verified $V_{ref}$ replaces the estimated one. If the sea-trial speed differs from the tank-test prediction, the attained EEDI is recomputed. Only after this final verification does the value become the certified attained EEDI. The result, the method, and the supporting data are recorded in the final EEDI Technical File, which stays aboard for the ship’s life and is reopened at any major conversion.

The output of all this is the International Energy Efficiency Certificate. The IEE Certificate, issued under Annex VI Regulation 6, records the verified attained EEDI, confirms it meets the required EEDI, and attests that a Ship Energy Efficiency Management Plan is aboard. The certificate is issued once and stays valid for the life of the ship unless a major conversion changes the EEDI, in which case the ship is re-verified and the certificate reissued. A ship can’t trade internationally without a valid IEE Certificate; port-state control checks for it alongside the IAPP certificate that covers the rest of Annex VI.

Major conversions and the EEDI obligation over a ship’s life

EEDI is set once at delivery, but it isn’t frozen forever. Annex VI Regulation 5 makes a major conversion a trigger for re-verification. A major conversion in this context is one that substantially alters the ship’s dimensions, carrying capacity, or engine power, the kind of change that would move the attained value. A jumboization that lengthens the hull and lifts deadweight, a re-engining that swaps the main engine for a different power or fuel, or a conversion that changes the ship’s type all force a fresh EEDI calculation and a reissued IEE Certificate. The required value the converted ship must meet is keyed to the date of the conversion, not the original keel-laying, so a 2014-built ship re-engined in 2025 faces the 2025 phase requirement.

This is also where EEDI touches the wider data regime. Since 1 January 2019 the IMO Data Collection System under Annex VI Regulation 27 has required ships of 5,000 GT and above to report annual fuel consumption, distance, and hours under way to their flag state, which forwards the data to the IMO ship-fuel-oil database. That dataset is operational, not design, so it feeds CII rather than EEDI directly, but it’s the same reporting backbone, and it’s how the IMO will check whether the design gains promised by EEDI translate into real fleet emissions over time. The attained EEDI of every ship, by contrast, is recorded in the IMO’s GISIS module so flag states and port-state-control officers can look it up.

How EEDI relates to EEXI, CII, and the wider Annex VI machinery

EEDI is the new-ship measure. It says nothing about the tens of thousands of ships already trading, and it says nothing about how a ship is actually run once it’s built. The IMO closed both gaps with two later measures that share EEDI’s DNA.

The Energy Efficiency Existing Ship Index extends the design-phase logic to the in-service fleet. EEXI uses the same attained-value formula as EEDI, with the same reference lines and a phase-style required value, but it applies to existing ships and was first assessed against the 2023 reporting cycle. Because most older ships can’t physically change their hull or engine, the usual route to EEXI compliance is power limitation: an engine power limitation or a shaft power limitation that caps the main-engine output, lowering the $P_{ME}$ term and the achievable $V_{ref}$ in the formula. A ship with an overridable limiter must log every time the override is used. EEXI is a one-time design-style check, computed once and certified, exactly like EEDI.

The Carbon Intensity Indicator covers what neither design index does: real operation. CII rates a ship from A to E each year on its actual measured CO2 emissions divided by its actual transport work over the year, using the fuel-consumption data already reported under the IMO Data Collection System. A ship can hold a strong EEDI and still earn a poor CII if it’s run badly, sails empty, or steams fast. EEDI is set in steel at delivery; CII moves every year with how the ship is operated, and a D or E rating triggers a corrective action plan in the SEEMP. The three measures stack: EEDI and EEXI fix the design ceiling, CII grades the operation, and all of them sit inside MARPOL Annex VI under the umbrella of the IMO GHG strategy targeting net-zero shipping emissions around 2050.

Alternative fuels and the limits of a tank-to-wake index

EEDI scores carbon at the engine, which makes the choice of fuel decisive for the attained value. LNG carries a 2.750 carbon factor against 3.114 for diesel and 3.206 for heavy fuel oil under MEPC.364(79), so a dual-fuel ship running on gas books a lower numerator and a better attained EEDI than an identical diesel ship, before any hull or propeller change. That arithmetic is part of why a large share of the post-2020 newbuild order book moved to LNG and methanol dual-fuel engines: the fuel switch helps clear the Phase 3 target on its own. Methanol and ammonia get their own factors as the IMO finalizes the lifecycle guidance, and a zero-carbon fuel burned at the engine would zero out that part of the numerator entirely.

The catch is the boundary. EEDI counts only the CO2 produced at the engine, the tank-to-wake emissions, and credits LNG for its lower combustion carbon. It doesn’t count the methane that slips through a low-pressure dual-fuel engine unburned, and methane has a global warming potential many times that of CO2 over a 20-year horizon, so a real-world LNG ship can have a worse climate footprint than its EEDI suggests. It also doesn’t count the well-to-tank emissions of producing the fuel, which for fossil LNG include liquefaction and for blue or grey hydrogen-derived fuels can be large. The EU’s FuelEU Maritime regulation, which began applying from 2025, takes the opposite approach and prices the full well-to-wake greenhouse-gas intensity of the fuel, so a fuel that looks clean under EEDI’s tank-to-wake boundary can rate worse under FuelEU. The two regimes pull in the same direction on decarbonization but measure different things, and a newbuild designed only to the EEDI boundary can find itself short on the EU side.

Limitations

EEDI binds only at the design stage, and only for new ships and major conversions. A ship delivered in 2012, the day before Phase 0 began, carries no EEDI obligation for its whole life unless it undergoes a major conversion that changes its index. The index says nothing about the existing fleet on its own; that gap is why the IMO had to add EEXI a decade later. So EEDI’s reach over the fleet’s total emissions is slow, gated by the pace of scrapping and newbuilding.

EEDI measures design, not operation, and the two diverge in practice. The attained value is computed at 75 percent MCR on the summer load draft in calm water, a single idealized point. A ship spends most of its life off that point: part-loaded, slow-steaming, in ballast, fighting weather. A good attained EEDI doesn’t guarantee low real-world emissions, which is the entire reason CII exists as a separate operational measure. Treating EEDI as a proxy for how green a ship actually is, is a category error.

The metric can be gamed at the edges, chiefly through the reference speed. Because $V_{ref}$ sits in the denominator and engine power scales with roughly the cube of speed, a designer who declares a conservative design speed can flatter the attained EEDI. Regulators tightened the sea-trial verification in MEPC.231(65) and its amendments partly to close this, requiring the measured speed-power curve rather than an optimistic claim, but the incentive to optimize for the test condition rather than the trade remains baked into any single-point index. The minimum-propulsion-power safeguard exists precisely because the same incentive can push installed power dangerously low.

EEDI also ignores emissions outside the tank-to-wake CO2 boundary. It doesn’t count methane slip from LNG and dual-fuel engines, which can erode the climate benefit of switching to gas; it doesn’t count the upstream, well-to-tank emissions of producing and delivering the fuel; and it doesn’t count nitrous oxide, black carbon, or other short-lived climate forcers. A well-to-wake or full-lifecycle accounting, which the EU’s FuelEU Maritime regulation and parts of the IMO’s evolving fuel-standard work move toward, would rate some EEDI-compliant fuel choices differently. Finally, the reference lines themselves are anchored to a 1999-to-2009 fleet baseline; as that baseline ages, the percentage-cut framing measures progress against a fixed historical point rather than against what’s technically achievable today.

See also

Calculators

• EEDI Attained Calculator
• EEDI Required Calculator
• EEDI Reference Line
• EEDI Phase Factor
• EEDI Innovative-Technology Credit
• EEDI Ice-class Correction (fj)
• EEDI Cubic-Capacity Correction (fcubic)
• EEDI Voluntary Structural Enhancement (fvse)

Related wiki articles

• What is EEXI? The Energy Efficiency eXisting-ship Index
• What is CII? Carbon Intensity Indicator explained
• MARPOL Annex VI
• SEEMP I, II and III
• IMO GHG strategy