Treatment of surfactant wastewater by foam separation: Combining the RSM method and WOA-BP neural network to explore optimal process conditions

Abstract

In the field of water treatment, the treatment of wastewater containing surfactants has been an important issue. The foam separation technique is often used to separate surfactants from water as a simple and efficient separation technique. We used a newly designed foam separation column to treat wastewater containing sodium dodecyl sulfonate (SDS) (0.5–2.0 mmol/L). The effects of initial SDS concentration (CSDS, i), operation time (t), foam height (H), gas velocity (V) and pH value (1−7) on the separation effect of SDS were investigated. The Box-Behnken design model was used to design the experimental group, and the process parameters of foam separation were optimized by combining response surface methodology (RSM) and WOA-BP neural network. The results showed that the prediction accuracy of the optimized BP neural network using the WOA algorithm was superior to that of the RSM model. The best operating parameters were: CSDS,i (0.88 mmol/L), t (28.13 min), H( 81.3 cm), V (1.03 L/min) and pH (3.5) and the recovery rate of SDS was up to 97.32%.