Applying quantitative structure-activity relationship (QSAR) models to extend the mixture toxicity prediction of scrubber water

Journal article, 2025

Marine organisms are constantly exposed to complex chemical mixtures from natural and anthropogenic sources. One source that has raised concerns is the discharge water from ships equipped with exhaust gas cleaning systems, commonly known as scrubbers. During operation, ships with scrubbers discharge large volumes of scrubber water, known to adversely affect marine organisms, into the environment. Scrubber water is highly acidic and contains a complex mixture of contaminants, including metals and polycyclic aromatic hydrocarbons (PAHs), at high concentrations. To assess the effect from these mixtures, critical values for individual mixture components can be determined from ecotoxicological studies and then compared to measured exposure concentrations. However, for several substances identified in scrubber water, for instance many alkylated PAHs, ecotoxicological studies are unavailable, preventing the determination of critical values. In this study, Quantitative Structure-Activity Relationship (QSAR) models have been used to amend and complement experimental data to estimate the mixture toxicity of scrubber water. Our results show that the combined predicted ecotoxicological response of an amended list of 50 substances measured in scrubber water from the substance groups metals (n = 10), PAHs (n = 16) and their alkylated derivatives (n = 24), still underestimates the response observed in whole effluent toxicity tests. This suggests that there are additional substances and/or synergistic effects in the scrubber water mixtures that contribute to the overall toxicity. Thus, to accurately describe the toxicity of scrubber water, measurements and toxicity assessments must extend far beyond the usual suspects of 16 PAHs and a limited selection of metals. Here, QSAR models and advanced chemical screening-based methods are valuable tools for identifying substances of concern.