Variations in toxicity of semi-coking wastewater treatment processes and their toxicity prediction

•The semi-coking WWTP showed a good performance in target contaminants removal.•Variation of biotoxicity in semi-coking wastewater treatment processes was studied.•Effective toxicity reduction was found in the coagulation and adsorption processes.•The CA model can be applied for toxicity prediction of semi-coking wastewater.

Chemical analyses and bioassays using Vibrio fischeri and Daphnia magna were conducted to evaluate comprehensively the variation of biotoxicity caused by contaminants in wastewater from a semi-coking wastewater treatment plant (WWTP). Pretreatment units (including an oil-water separator, a phenols extraction tower, an ammonia stripping tower, and a regulation tank) followed by treatment units (including anaerobic-oxic treatment units, coagulation-sedimentation treatment units, and an active carbon adsorption column) were employed in the semi-coking WWTP. Five benzenes, 11 phenols, and five polycyclic aromatic hydrocarbons (PAHs) were investigated as the dominant contaminants in semi-coking wastewater. Because of residual extractant, the phenols extraction process increased acute toxicity to V. fischeri and immobilization and lethal toxicity to D. magna. The acute toxicity of pretreated wastewater to V. fischeri was still higher than that of raw semi-coking wastewater, even though 90.0% of benzenes, 94.8% of phenols, and 81.0% of PAHs were removed. After wastewater pretreatment, phenols and PAHs were mainly removed by anaerobic-oxic and coagulation-sedimentation treatment processes respectively, and a subsequent active carbon adsorption process further reduced the concentrations of all target chemicals to below detection limits. An effective biotoxicity reduction was found during the coagulation-sedimentation and active carbon adsorption treatment processes. The concentration addition model can be applied for toxicity prediction of wastewater from the semi-coking WWTP. The deviation between the measured and predicted toxicity results may result from the effects of compounds not detectable by instrumental analyses, the synergistic effect of detected contaminants, or possible transformation products.