Opportunity NO_x: How to fix the IMO’s backwards air pollution regulation

Nitrogen oxides (NO_x) are air pollutants emitted when internal combustion engines burn fuel. NO_x emissions contribute to the formation of smog and acid rain and, more broadly, contribute to human health problems and environmental damage. The United Nations International Maritime Organization (IMO) has been regulating NO_x emissions from marine engines for decades, but new research from the ICCT and partners has revealed that ships with newer engines are emitting more NO_x than those with older engines.

We found that Tier II engines, which were built between 2011–2015, emitted an average of 10.9 grams of NO_x per kilowatt-hour (g/kWh), compared with 9.8 g/kWh for older Tier I engines, which were built between 2000–2010. This difference was statistically significant (T-test, p <0.01). These statistics were derived from 311 measurements of Tier I engines and 124 measurements of Tier II engines, covering all engine speeds (rpm).

It’s important to note that the IMO differentiates its NO_x regulations by rpm. The higher the engine rpm, the lower the allowable NO_x limit. When we looked at the data, we saw something interesting: newer Tier II engines tended to have lower rpm than older Tier I engines, as shown in Figure 1. The lower the rpm, the larger the engine, and the higher the allowable NO_x limit, with the highest limits applying to engines <130 rpm.

In Figure 1, the left-most cluster of dots are the NO_x measurements from engines <130 rpm. It’s reasonable to wonder whether the higher NO_x emissions we observed for Tier II engines are simply because these are mainly larger, lower-rpm engines. To investigate this question, we compared measured NO_x emissions rates for Tier I and Tier II engines <130 rpm.

Figure 1. NO_x emission rates (g/kWh) by engine rpm and tier, compared with weighted regulatory limits.

Figure 2 shows a boxplot of NO_x emission rates for ships with Tier I and Tier II engines <130 rpm. We had 121 measurements of ships with Tier I engines <130 rpm and 100 measurements for ships with Tier II engines <130 rpm. On average, Tier I emitted 10.7 g/kWh, compared with 12.0 g/kWh for Tier II. This difference was statistically significant (T-test, p <0.01). By comparing results from the same rpm range, we clearly determined that the increase in NO_x emissions for ships with Tier II engines isn’t just because the newer engines tend to be larger and slower.

Note: x shows mean; horizontal bars inside the boxes show median; boxes show interquartile range; whiskers show minimum and maximum (excluding outliers); dots show outliers.
Figure 2. Boxplot showing the distribution of observed NO_x emission rates for ships with Tier I and Tier II engines with engine speeds below 130 rpm.

Obviously, newer engines shouldn’t emit more air pollution than older engines, especially when Tier II engines are supposed to emit 15–20% less NO_x than Tier I engines, according to IMO regulations. In our study, we pointed out that engines are certified on a test cycle that weights emissions at higher engine loads more than those at lower engine loads. In reality, ships tend to operate at lower engine loads, when NO_x emission rates are higher. The highest NO_x emission rates were observed when ships were operating at below 25% main engine load, averaging 11.6 g/kWh for ships with older Tier I engines (all rpms) and 13.5 g/kWh for ships with newer Tier II engines (all rpms). Emissions below 25% engine load are omitted from the NO_x engine certification scheme for most engines, meaning they are essentially unregulated.

To fix IMO’s NO_x regulation, member states should add an additional test point below 25% engine load. Data from our NO_x study shows that ships with Tier I or Tier II engines are spending approximately 5% of their time operating below 5% engine load and 10% of their time operating below 15% engine load. To split the difference, a new test point at 10% engine load could help characterize low-load emissions. In addition to this new test point, we recommend that the IMO consider establishing not-to-exceed (NTE) standards that prevent excess pollution across all engine loads, even those outside the test points. These limits should apply to both new and existing engines because waiting for the shipping fleet to turnover (the average ship is more than 22 years old) risks long delays in providing relief to coastal communities that are most impacted by shipping’s air pollution.

Properly regulating NO_x emissions will continue to be important, even as the sector works to implement its decarbonization strategy (which we analyzed here) because any fuel that is combusted in a marine engine – hydrogen, ammonia, methanol, etc. – produces NO_x. The next opportunity for IMO delegates to consider how to revise NO_x regulations comes at the 11th session of IMO’s Pollution Prevention and Response (PPR) subcommittee meeting in February 2024.

Member states and international organizations have been invited to provide information on in-service engine NO_x emissions measurement campaigns, including findings from recent studies. That’s a good start, and our study results will be included in a PPR submission, but IMO delegates should go farther. At PPR 11, member states and international organizations should recommend that the IMO’s Marine Environment Protection Committee establish a new work output focused on revising the NO_x standards for new and existing ships, with a particular focus on reducing NO_x emissions at low engine loads. Opportunity knocks, and it’s time to answer.