10.1021/ie5010898.s001 Mingsen Liao Mingsen Liao Yuehong Zhao Yuehong Zhao Pengge Ning Pengge Ning Hongbin Cao Hongbin Cao Hao Wen Hao Wen Yi Zhang Yi Zhang Optimal Design of Solvent Blend and Its Application in Coking Wastewater Treatment Process American Chemical Society 2014 NLP model case study blend exhibits Optimal Design MIBK concentration MIBK loss nonlinear programming coking wastewater treatment candidate diluters makeup cost distribution coefficient extraction performance molar fraction Coking Wastewater Treatment ProcessOne tar removal phenol 2014-10-01 00:00:00 Journal contribution https://acs.figshare.com/articles/journal_contribution/Optimal_Design_of_Solvent_Blend_and_Its_Application_in_Coking_Wastewater_Treatment_Process/2250724 One of the key steps of coking wastewater treatment is phenolic and tar removal via extraction. However, the high loss of the extractant, i.e., methyl isobutyl ketone (MIBK), leads to the high cost of the process. The adoption of a novel solvent or solvent blend is considered as an efficient way to address this problem. In this paper, seven solvents (benzene, toluene, m-xylene, ethylbenzene, 1, 3, 5-trimethylbenze, cyclohexane, and octanol), selected as candidate diluters for MIBK according to operating requirements, are studied with a nonlinear programming (NLP) model based on ideal counter-current extraction. The results, verified with experiments, suggest toluene is the most promising candidate. Further investigation of this solvent blend reveals that both <i>D</i><sub>blend</sub> (the distribution coefficient of phenol between solvent blend and water) and <i>m</i><sub>MIBK</sub> (the MIBK concentration in raffinate) increase with <i>x</i><sub>MIBK</sub> (the molar fraction of MIBK in blend). The trade-off between the extraction performance and MIBK loss recommends the blend with <i>x</i><sub>MIBK</sub> = 0.05 as extractant for coking wastewater treatment. An industrial process consisting of extraction, back stripping, distillation, and mixer is presented. A corresponding NLP model is established for its operating optimization. To improve the accuracy, the representatives of typical phenolics and tar in wastewater (2,4 dimethyl phenol, m-xylene, and quinolone) are also considered in addition to phenol. The case study indicates that the blend exhibits economic advantage over pure MIBK with a makeup cost of 11.15 ¥/t, much less than the 185.15 ¥/t in the case of MIBK.