**4. Discussion**

The reports of ICCT and Sphera present average methane slip values of 5.5 g/kWh and 3.8 g/kWh for LPDF engines, respectively [14,15]. However, the methane emissions from actual LNG vessels have not ye<sup>t</sup> been widely studied. Anderson et al. [8] conducted onboard studies on an LNG ship resulting in total hydrocarbon emissions of 1.1–6.7 g/kWh (depending on engine load). Ushakov et al. [12] report average methane emission of 5.26 g/kWh based on on-board measurements. Peng et al. [6] also did on-board studies and reported methane levels of 3.7–25.5 g/kWh, the lowest level corresponding to highest engine load and the highest methane level corresponding to engine load of 25%. They also publish the actual methane concentration of 1000 ppm at 90% engine load and higher concentrations at lower engine loads. Lehtoranta et al. [5] reported methane levels from one medium speed engine (dual fuel, natural gas) at two load modes of 85% and 40% resulting in 5.6 g/kWh and 13.8 g/kWh, respectively. These levels correspond to methane concentrations of 1800 ppm (at 85% load) and 3750 ppm (at 40% load) (Table 2). Wei and Geng [30] review also reports total hydrocarbon concentrations above 2000 ppm from natural gas/diesel dual fuel combustion. Comparing these published results, there is a large variation in the methane levels reported, depending, most probably, on different engines (e.g., engine model, size, age) used as well as on engine loadings. Most of the values reported are in g/kWh while there are few publications reporting in concentrations (ppm) also. By comparing these values, it can be concluded that all the methane slips reported are in the order of 1000 ppm or higher. Furthermore, comparing this to the present study, the methane level (1000 ppm and 1500 ppm) in the present study was in the same order of magnitude. However, we used a smaller engine to investigate the performance of MOC but with a target to have similar exhaust gas composition to medium speed DF engines operating with NG as the main fuel. Therefore, we consider the results of the present study to be interpreted as indications of the possibilities to use MOC in LNG vessels.

In the present study, MOC was found to decrease the methane emission by 70–80% at the exhaust temperature of 550 ◦C. This efficiency, however, decreased significantly within time, even with only 0.5 ppm SO2 in the exhaust. Regeneration was done once a day and was found to recover the efficiency. This regeneration was done in nearly stoichiometric conditions, the very low O2 enabling SO2 release from the MOC while also the temperature increase observed during the regeneration phase might influence the SO2 release. In addition, the H2 might play a role here too, since H2 was found in the exhaust in the regeneration phase and has also been used for methane catalyst regeneration purposes in earlier studies [26,27].

Overall, it seems that the temperature is in significant role, both in methane oxidation efficiency as well as in regeneration. One regeneration performed in the present study resulted in higher exhaust temperature downstream of the catalyst and released higher amount of SO2 compared to other regenerations.

In addition to MOC, a SOx trap was studied in connection to MOC, for the first time (to the authors' knowledge). The SOx trap was shown to protect the MOC against sulfur poisoning to some extent, since tripling the exhaust SO2 level (from 0.5 ppm to 1.5 ppm) was found to have, in practice, no effect on MOC performance when trap was used upstream the MOC.

Other hydrocarbon emissions (ethane, ethylene) and carbon monoxide are easier to oxidize and were found to be nearly totally diminished by the MOC over the whole ageing test. The NG in the present study was high in methane content, while the fuel gas composition from different suppliers can have lower methane content (see e.g., [12]) and higher share of ethane and heavier hydrocarbons favoring better performance of MOC.

Formaldehyde, HCHO, is defined as carcinogenic substance and contribute also to other severe health effects, such as asthma, see, e.g., [31,32]. Therefore it is important to also consider the HCHO increase by LNG use (e.g., [6]), especially in coastal areas. In the present study, formaldehyde measurements were employed and showed that the

MOC very effectively (by 95–99%) reduced the formaldehyde emissions and practically no decrease was found in the formaldehyde oxidation efficiency during the catalyst ageing studies. This indicates that formaldehyde emissions from NG engines can be diminished with the MOC employment.

The results of the present study give indication of the possible use of MOC in LNG ships to control methane slip emissions. However, the regeneration process in real sized lean-burn marine engine is an issue that needs to be solved. Optimization of the regeneration interval and duration depending on the actual target of application is needed as well. Further catalyst development, regarding the efficiency and sizing, is to be done to have the best possible catalyst suitable to be installed at high-temperature conditions in the exhaust line.

Currently, LNG is a viable marine fuel deployed to substantially reduce pollutant emissions (NOx, SOx and particles) from ships. The use of LNG has therefore a notable effect on air quality and human health. In addition, LNG can provide reduction in GHG emissions if methane slip is controlled. This methane slip challenge needs to be solved to maximize LNG's potential to contribute to climate neutrality. The present study shows that MOC can play a role here. As it is an after-treatment system, it has potential both in new vessels as well as to retrofit to existing vessels. However, further studies are needed to solve the optimized solution for MOC and performance on different engine loadings as well in transient loading relevant in vessel operation.

In the long term, the transition from fossil NG to renewable NG is needed. This means also that further studies on, e.g., biogas (LBG) is needed to be able to develop suitable systems for biogas emissions since, e.g., possible biogas impurities may play a major role in the catalyst performance.

**Author Contributions:** Conceptualization, K.L., H.V., K.K. and T.M.; methodology, K.L., P.K., H.V.; investigation, K.L., P.K., H.V., K.K. and T.M.; original draft preparation, K.L.; writing—review and editing, K.L., P.K., H.V., K.K. and T.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was part of INTENS-research project, funded by Business Finland and several Finnish companies.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
