and alkylnaphthalenes were lower in the CL-discharge water (sum of averages: 24.4 %) than in the fuels (51.8 %), presumably as a result of the volatility of naphthalenes. Hence, Phe and alkylphenanthrene frac- tions were higher in the CL-discharge water (49.1 %) than in the fuels (17.2 %). This effect is accentuated in S3 which had a lower specific water flowrate (i.e. the recirculated water had a higher residence time; Table 2) than S1 and S2 in CL-operation. The similarity in the PAH patterns between fuel and water samples confirms a predominantly petrogenic nature of the PAH in the discharge water samples. Some results for S3 contrasted with the similarities be- tween S1 and S2, largely explainable by the fuel quality. Another approach we followed to compare the PAH content in the fuel with the PAH concentrations in the discharge water was to consider the emission factors. For the fuels, the PAH concentrations were multi- plied by the specific fuel consumption (kg/MWh) of the ships corre- sponding to the time of the generation of the sampled water fromOL and CL operation (Table 2). The mass of PAH (from the fuel) flowing into the engine per energy output was between 0.87 and 5.17 kg/MWh in S1–3 for the P 71 PAH. These results were used (as divisor) to calculate a ratio with the emission factors from the water discharges. Assuming 100 % petrogenic origin of PAH, the ratios indicate the proportion of PAH of the fuel that ends up in the discharge water. The ratios P 71 PAH in OL (0.25–1.74  10–3) were one order of magnitude higher than in CL (0.09–2.83  10–4). The highest ratio, which was in OL of S1, indicate that 0.17 % of the P 71 PAH in the fuel is found in the discharge water; the remaining PAH (>99.8 %) are either burnt in the engine or emitted into the atmosphere. If the PAH patterns (percentage distribution) of the fuel samples are subtracted from the PAH patterns (percentage distribution) of the cor- responding discharge water samples for each ship (Suppl. Inform. Fig. S7), the loss of acenapthylene (Acy) and Ant in all samples is observed probably due to low stability during the scrubbing process. Lowest Ant and Acy concentrations among the LMW PAH were also observed by Hermansson et al. (2021), Teuchies et al. (2020) and Ush- akov et al. (2020). An enrichment is observed for Nap (in OL), Ace and Phe (in OL and CL). Notably, 3-methylphenanthrene concentrations are lower in all fuels and higher in all discharge water samples whereas the opposite is true for 9-methylphenanthrene (Fig. 7), which is in line with the lower stability of 9-methylphenanthrene vs. 3-methylphenanthrene as used in the methylphenanthrene indices in coalification studies (Radke et al., 1982). 3.3. Concentrations of PAXH Heterocyclic compounds containing sulfur, oxygen or nitrogen were identified both in the fuels and in the scrubber waters. Some of their alkylated derivatives were also detected, especially in the fuels, but a lack of reference compounds precluded an extensive qualitative and quantitative analysis of these species. As expected, higher concentra- tions were found in the scrubber waters derived from the CL-operation, except for S4. The concentrations are low (Suppl. Inform. S4) and actually the sum of the concentrations for the seven two- and three-ring unsubstituted heterocycles with the three heteroatoms mentioned were on the same order as that of Phe alone. 3.4. Non-target screening of PAC PAC-specific screening of the discharge water samples of the four ships using laser ionization largely confirmed the results of Garcia- Gomez and co-workers (2023) which is exemplarily shown in Suppl. Inform. Table S8. PAH for which standards were available are marked in bold and all other assignments were made according to the mass and plausibility of retention time. Mostly alkylated 2- and 3-ring PAH and less alkylated 4-ring PAHwere detected (in addition to the unsubstituted compounds Nap, Fl, Phe, Pyr and Chr). 5- and 6-ring PAH were hardly found and HMW PAH in discharge water sample S3 showed a higher presence than in the other samples which is in line with findings for target PAH in S3 discharge water. Alkylbiphenyls, dibenzofurans and dibenzothiophenes (in addition to unsubstituted substances) were also confirmed. For the first time, C5- and C6-naphthalenes and C3- Fig. 7. Relative mass fraction (%) of LMW PAH in the fuel and OL dissolved fraction samples of ships 1–3.P71 PAH are additionally given in the box: PAH in fuel/ PAH in OL-discharge of each ship. C. Achten et al. Marine Pollution Bulletin 208 (2024) 116790 10