*Test Program*

The experiments were conducted over a 190 h' ageing at the selected driving mode. The driving mode selection was based on the exhaust gas composition with an aim to produce as similar exhaust gas as possible that would be emitted from LNG engine. The engine out exhaust gas composition on the selected driving mode is presented in Table 1 and shows that the NOx level was approx. 225 ppm and CO 550 ppm, while hydrocarbons constituted of methane with near 1000 ppm level, ethane 15 ppm, and ethylene 30 ppm. In addition, formaldehyde (HCHO) was found in the exhaust with a level of approx. 55 ppm. For comparison and as a reference we have added the exhaust gas composition measured form one medium speed dual fuel marine engine run with NG to Table 2. These were earlier presented in g/kWh [5] and now in concentrations at two engine loads of 85% and 40%. Comparing these values to the values of the present study's selected driving mode (Table 1 "Ageing") we see that these are in the same order of magnitude.

**Table 1.** Engine out exhaust gas composition (as measured in dry gas). In "Ageing"-mode hydrocarbons are measured by GC and others by Horiba, while in "Regeneration"-mode O2 and H2 were measured by GC and others by FTIR. HCHO was measured with FTIR.


**Table 2.** Engine out exhaust gas composition from one medium speed dual fuel marine engine run with natural gas as the main fuel [5].


To ensure that the starting point for the experiments was as similar to possible for all catalyst reactors, they were preconditioned by ageing for 48 h in the selected driving mode with an exhaust gas temperature of 550 ◦C and exhaust flow of 60 kg/h. After the preconditioning, the initial performance of the MOC was studied by emission measurements. The same driving mode was used. The methane level in this driving mode was near 1000 ppm (Table 1) and in addition, one higher methane level was included in the initial performance studies to see if the methane level influences the performance. This approx. 1500 ppm methane level was achieved in the same driving mode by adding methane to the exhaust prior to the catalyst.

After the initial performance studies, the ageing tests were conducted. Additional SO2 was fed into the exhaust (see Figure 1) in part of the ageing tests, while one test was also conducted without any additional SO2. The added SO2 contributed to a 1 ppm increase in the exhaust gas while any sulfur from the natural gas and lubricating oil, led to a SO2 level of approximately 0.5 ppm in the exhaust gas. This means SO2 level of 0.5 ppm without any additional SO2 and SO2 level of 1.5 ppm with the added SO2 in the exhaust. The SO2 level of 1.5 ppm was confirmed by sampling exhaust gas to a sample bag and analyzing by gas chromatography prior to the ageing tests. In the following ageing tests, the SO2 feed was then kept constant.

Altogether, three similar experimental campaigns were conducted. Two ageing tests were done with MOC only, meaning one ageing test with additional SO2 and the other one without SO2 addition. The third ageing test was conducted with SOx trap installed in front of the MOC (including the additional SO2 in the exhaust gas).

During the ageing test, the engine was running without stops and once a day regeneration was done by turning the engine to stoichiometric condition for 5 min time. In Table 1, also the engine out emissions at the regeneration mode are presented. This shows that in addition to the targeted O2 decrease, other emission components' levels changed as well. The NOx increased to 2400 ppm and CO to above 9000 ppm in the stoichiometric mode while methane level decreased to 860 ppm. In addition, H2 with a level of 0.6% was found in the exhaust when running on the nearly stoichiometric mode.
