Experimental Optimization and Predictive Modeling of Ultrasonic Remediation of Used Cooking Oil–Contaminated Sand

Used cooking oil (UCO) released into the environment can contaminate sand and create persistent remediation challenges. This study evaluates ultrasonic treatment as a practical approach for removing UCO from oil contaminated sand and develops a predictive model for estimating removal efficiency. A central composite experimental design comprising 52 runs was employed to investigate the effects of treatment time, ultrasonic power, aqueous medium, and oil–sand ratio. Residual oil concentrations before and after treatment were quantified using ultraviolet–visible (UV–Vis) spectroscopy, while thermogravimetric analysis (TGA) was used to support the observed removal trends. The results demonstrated that removal efficiency was strongly influenced by the combined operating conditions. The highest removal efficiency of 41.4% was achieved at 240 W, 15 min, an oil–sand ratio of 1:0.5, and saline water, whereas the lowest efficiency of 1.3% occurred at 120 W, 5 min, a ratio of 1:1.5, and tap water. Increased treatment time, ultrasonic power, and salinity generally enhanced oil removal, while the oil–sand ratio showed the strongest influence in the regression analysis. A linear regression model was subsequently developed to predict removal efficiency within the investigated operating range, achieving a root mean square error (RMSE) of 2.73. These findings demonstrate the potential of ultrasonic treatment as a controllable remediation technique for UCO-contaminated sand and provide a quantitative basis for further process optimization and scale-up.