[Updated] Choose wisely: IMO’s carbon intensity target could be the difference between rising or falling shipping emissions this decade

Strategies Tracking progress
Emissions modeling

The International Maritime Organization’s (IMO) initial greenhouse gas (GHG) strategy calls for emissions to peak as soon as possible, to fall at least 50% by 2050 compared to 2008, and for at least a 40% reduction in the carbon intensity of international shipping by 2030 relative to 2008. On that last part, the IMO hasn’t yet defined “carbon intensity,” and at the 8th meeting of IMO’s GHG intersessional working group next week, member states will pick one of two competing ways to measure it: the energy efficiency operational indicator (EEOI) or the annual efficiency ratio (AER). Then they’ll need to agree on how quickly to reduce the carbon intensity of the fleet.

Member states should choose wisely. IMO’s GHG strategy also aims to pursue efforts to phase out emissions on a pathway consistent with Paris Agreement temperature goals, which would include efforts to limit warming to no more than 1.5°C. In this context the carbon intensity choice matters a great deal, because our analysis shows that achieving the bare minimum 40% reduction based on EEOI could allow emissions to grow 22% this decade, which is even worse than business as usual (BAU). Achieving a 40% reduction based on AER still allows emissions to grow more than 5%; this is better than BAU, but higher than the 2008 peak year.

Keep in mind that I’m assuming all ships reduce their carbon intensity. In reality, only a portion of the fleet will be covered by the policy, due to proposed exemptions and correction factors, as well as data limitations. Ships reporting to the IMO’s Data Collection System (DCS), which covers ships greater than 5,000 gross tonnes, emitted 614 million tonnes (Mt) of carbon dioxide (CO2) in 2019, which is only about two-thirds of the 919 Mt CO2 emitted by international shipping in 2018, as estimated in the Fourth IMO Greenhouse Gas Study. The DCS will probably be used to ensure compliance with the regulation. If the policy ends up covering at best two-thirds of the emissions, then policymakers need to err on the side of higher ambition to ensure that emissions fall this decade, instead of rising.

As shown in the figure below, greater emissions reductions can be achieved if IMO chooses annual reductions in carbon intensity that are aligned with keeping emissions below the 2008 peak, -3% per year, achieving its 2050 emissions target, -4% per year, or limiting warming to 1.5°C, -6% to -7% per year. The table at the end of this post has details of the methodology. So, you can see that it’s the choice of carbon intensity reduction rate that really matters here. No matter which way IMO decides to measure its progress to its 2030 target, it has already decided that all ships will need to reduce their carbon intensity from today’s levels, even if they’ve already achieved a 40% reduction in carbon intensity relative to 2008. This prevents backsliding. They’ve also decided that ships will comply with the carbon intensity regulation by reducing their AER each year. This is because IMO already collects data that can be used to estimate a ship’s AER, but not it’s EEOI; more on that later. This implies that a reduction in AER reduces EEOI by the same amount. It’s a simplifying assumption that works well enough to estimate how reducing carbon intensity will affect absolute emissions between now and 2030, as I’ve done in the figure.

Figure. International shipping CO2 emissions (million tonnes) from 2018 to 2030 and associated annual reductions in carbon intensity (gCO2/dwt-nm) required to achieve various policy objectives, assuming full compliance and no exemptions or correction factors for any ship.

What makes the two ways to measure carbon intensity so different? The EEOI is a “demand-based” efficiency metric, a measure of the real-world carbon intensity of the fleet because it estimates how much CO2 was emitted to transport 1 tonne of cargo 1 nautical mile (gCO2/t-nm). The AER is a “supply-based” efficiency metric, a measure of the theoretical carbon intensity of the fleet because it divides the amount of CO2 a ship emits by its cargo carrying capacity (deadweight tonnes, or dwt), no matter how full the ship is, and then again by the nautical miles the ship traveled in a year (gCO2/dwt-nm). You might be thinking: If we have the choice, why wouldn’t we go with the EEOI over the AER? Two reasons. First, the IMO’s DCS contains all the information needed to calculate AER for each ship, but it’s missing actual cargo data; that means the IMO would need to modify its data reporting requirements to have an official estimate of international shipping’s EEOI. This process could take years and there’s no guarantee that IMO member states will ever agree to force ships to report cargo data. Second, it turns out that setting the carbon intensity target based on EEOI is less ambitious than setting it based on AER.

Under the EEOI metric, ships had already reduced their carbon intensity 32% relative to 2008 as of 2018, according to the Fourth IMO Greenhouse Gas Study. That means we only need a small annual fleetwide efficiency improvement from 2018 to 2030 to get to the IMO’s minimum 40% reduction target. Reducing the carbon intensity is great, but because demand for shipping is growing faster than efficiency is improving, we can expect total emissions from the global fleet of ships to continue to grow. Indeed, one of the business-as-usual projection scenarios in the Fourth IMO Greenhouse Gas Study, which we use in this analysis, predicts a 16% increase in emissions between 2018 and 2030, despite a 17% reduction in carbon intensity over that time. This is equivalent to a 43% reduction from 2008 and means that aiming for only a 40% reduction based on EEOI is actually a step backward.

If we instead try to achieve the minimum 40% reduction in carbon intensity based on AER, we still see emissions grow, just not as much. Under the AER metric, shipping’s carbon intensity in 2018 was 22% better than 2008, according to the Fourth IMO Greenhouse Gas study, and shipping’s carbon intensity has to fall about 2% annually starting in 2019 to achieve the minimum 40% reduction from 2008 levels by 2030. By choosing AER, instead of rising 22%, we expect emissions to grow 5% by 2030 compared to 2018. So aiming for a 40% reduction in AER doesn’t even peak emissions.

Achieving a 40% reduction on either metric allows emissions to continue to grow this decade, which makes it harder to achieve IMO’s 2050 target, and growing emissions are wholly incompatible with the Paris Agreement temperature goals.

So, how to get on a 1.5°C-compatible trajectory? Instead of requiring the bare minimum to achieve the 2030 carbon intensity target, IMO member states could do much more to align with the organization’s goals. Let’s sum things up.

To keep emissions below the 2008 peak:

  • Require at least a 3% annual reduction in carbon intensity per year, equal to a 33% reduction from 2019–2030.

To be aligned with IMO’s 2050 target:

  • Require at least a 4% reduction in carbon intensity per year, equal to a 44% reduction from 2019–2030.

To be aligned with a 1.5°C pathway:

  • Require at least a 6%–7% reduction in carbon intensity per year, equal to a 66% to 77% reduction from 2019–2030.

It’s up to IMO delegates to choose at the intersessional meeting, which will be held virtually from 24–28 May 2021. They’ve been charged with putting the international shipping sector on a pathway that helps society achieve its climate goals, and now’s their chance to make a difference. It’s important that they choose wisely.

Table. Details of the methodology used to generate the figure.
Trajectory Source or methodology
2008 baseline International shipping emissions, based on the Third IMO Greenhouse Gas Study.
BAU The Fourth IMO Greenhouse Gas Study’s SSP2_RCP2.6_L business-as-usual scenario expects a 16% increase in total shipping emissions from 2018 to 2030. We’ve estimated BAU international shipping emissions by assuming they also increase 16% from 2018 levels by 2030, resulting in 1,062 Mt of international shipping emissions (vessel-based approach) in 2030. Under this scenario, demand for shipping grows 39% from 2018 to 2030, while emissions grow only 16%. This implies a 17% reduction in carbon intensity from 2018 to 2030. This is contrary to the 10% reduction assumption reported in Table 41 of Annex K of the Fourth IMO Greenhouse Gas Study, which appears to be an error.
EEXI Assumes that the Energy Efficiency Existing Ship Index (EEXI) results in about a 1% reduction in BAU 2030 emissions, based on this ICCT research.
-40% EEOI aligned Assumes ships just barely achieve a 40% reduction in EEOI compared to 2008 levels by 2030. EEOI was 17.1 gCO2/t-nm in 2008 and 11.67 in 2018, according to the Fourth IMO GHG Study. By 2030, an additional 17% improvement from 2018 levels is projected under the SSP2_RCP2.6_L scenario, which equals 9.68 gCO2/t-nm in 2030, a 43% reduction from 2008. A 40% reduction from 2008 would equal 10.26 gCO2/t-nm, which is worse than BAU carbon intensity. Given that BAU emissions are expected to be 1,062 Mt CO2 in 2030, we can estimate 2030 emissions under a 40% reduction in EEOI as follows: 1,062 * (10.26/9.68) = 1,125 Mt CO2.
-40% AER aligned Assumes that the equivalent reduction in EEOI carbon intensity to just barely achieve a 40% reduction in AER by 2030 compared to 2008 is twice as much as needed to achieve a 40% reduction in EEOI. It also assumes that from 2019 onward, a percent reduction in AER is results in an equivalent percent reduction in EEOI. This is based on modeling done by the coordinators of the IMO Correspondence Group on the development of technical guidelines on carbon intensity reduction. This results in an equivalent EEOI of 8.85 gCO2/t-nm in 2030 to achieve a 40% reduction in AER (equivalent to a 48% reduction in EEOI relative to 2008). As mentioned before, BAU carbon intensity is predicted to be 9.68 gCO2/t-nm in 2030. Given that BAU emissions are expected to be 1,062 Mt CO2 in 2030, we can estimate 2030 emissions under a 40% reduction in AER as follows: 1,062 * (8.85/9.68) = 970 Mt CO2.
Keep below 2008 peak Assumes carbon intensity falls 3% per year which results in a 2030 carbon intensity of 7.72 gCO2/t-nm. Under full compliance, this results in 847 Mt CO2 in 2030, as follows: 1,062 * (7.72/9.68) = 847 Mt CO2.
IMO 2050 aligned Assumes international shipping emissions are 50% lower in 2050 compared to 2008 and draws a straight-line trajectory from 2008 to 2030.
1.5°C aligned (ICCT) Based on the IPCC Special Report on Global Warming of 1.5 °C, to have a 67% chance of keeping warming below 1.5°C, the world’s remaining carbon budget from 2018 onward is 420 gigatonnes (Gt) of CO2. On average, between 2012 and 2018, international shipping accounted for 2.47% of anthropogenic CO2 emissions, according to the Fourth IMO Greenhouse Gas Study; assuming a similar share going forward, this suggests that shipping has about 10,362 Mt of CO2 budget remaining from 2018 onward. Emitting only that amount would mean zero emissions by 2039 and that emissions reductions start in 2019; this allows for 414 Mt of CO2 in 2030. If emissions are not yet decreasing, fewer emissions will be allowed in 2030 within the carbon budget, and it would require faster improvements in annual carbon intensity.
1.5°C aligned (CSC & PE) Based on the 1.5°C aligned pathway presented by Clean Shipping Coalition and Pacific Environment in Figure 1 of document ISWG-GHG 7/2/12. Note that their pathway assumes that global as opposed to international shipping emissions fall to zero no later than 2035 and that absolute emissions do not begin to fall until 2023.

This blog was revised on 21 May, 2021 to correct an apparent error in the business as usual carbon intensity improvement from 2018 to 2030 reported by the Fourth IMO Greenhouse Gas Study, which we estimate should be a 17% improvement from 2018 to 2030 (43% improvement relative to 2008), rather than a 10% improvement from 2018 to 2030 (39% improvement relative to 2008). Please see the “BAU” row in the table at the end of the blog post for more information. The previous version can be found here.