Optimisation of dissolved air flotation (DAF) for separating industrial mineral oil from water

Industrial mineral oil wastewater from oil refineries and petrochemical processing poses a major environmental concern. Effluents from these processes is usually poor as it is heavily polluted, thus have high chemical oxygen demand (COD), soap oil and grease (SOG), turbidity, total suspended solids (TSS) amongst others. This wastewater, if discharge without treatment, causes severe pollution, oxygen depletion, and imbalanced ecosystem and human health risks. The main aim of this research was to modify, optimise and evaluate the performance of a continuous process using dissolved air flotation (DAF) pilot to treat wastewater from a local South African oil refinery wastewater treatment plant (WWTP) with the benefit of recovery of the oil from the wastewater. The study evaluated the feasibility of using different acids and coagulants. One factor at time (OFAT) approach was used on the DAF jar tester to identify the most important variables that affects the DAF treatability performance. The factors considered were; pH, flotation time, coagulant dosage, air to water ratio and air saturated pressure. The ranges considered for the factors were pH (4−6), flotation time (5−15 minutes), coagulant dosage (10−50 mg/L), air to water ratio (5–15%) and air saturated pressure (300–500 kPa). The key process operating parameters obtained from the OFAT were optimised using the Box Behnken design (BBD) adapted from response surface methodology (RSM). The BBD used had three levels, three factors and five centre points. This was employed to establish the relationship that existed between the water quality (contaminants) and the key interacting factors of the DAF jar tester, thus employing the most applicable combination of the factors on a continuous DAF pilot plant. The study was configured into two; Acid – Coagulation-DAF (pre-treatment) and Acid –DAF – Coagulation (post treatment). Three acids were investigated for their efficiency in the pre- treatment step, while four cationic inorganic coagulants and three polymeric organic coagulants were used both for the pre and post treatments. The OFAT experiments resulted in more than 75% removal efficiency of COD, SOG, TSS and turbidity. The removal efficiency was obtained at the following optimum values of pH 5, flotation time of 15 minutes at a coagulant dosage of 50 mg/L and an air to water ratio of 10% and finally, air saturated pressure was 350 kPa. On the other hand, BBD results showed 85% treatability performance at a lower coagulant dosage (30–45 mg/L), moderate air saturator pressure (300–425 kPa), and air-water ratio (8–12%) on the batch scale. While on the continuous process, the optimum coagulant dosage was around 100–180 mg/L. From the BBD results, the interacted factors for consideration were the air saturated pressure and coagulant dosage. These factors enhanced process control. The validation of all the response quadratic models were in good standing with the analysis of variance (ANOVA). The experimental results and the predicted models results agreed at 95% confidence level, finally, the models were significant and verified. Comparative studies of the pre and post treatment showed that 1 M H3PO4 was the most effective, economical and environmentally friendly acid to be used for both processes. Two cationic inorganic (alum and ferric chloride) and two polymeric organic (Z553D-PAC and Zetag32-FS/A50) coagulants were found to be effective with remarkable performance to destabilise and neutralise the oil droplets to coalesce larger flocs to enhance the oil-water separation. Far and above, the cationic inorganic coagulants were more cost effective than the polymeric organic coagulants, even though, the inorganic coagulants were cheaper they had higher conductivity (salts), thus raising environmental concerns. In conclusion, the pre-treatment of the DAF process yielded more recovery of water and oil, and hence this step was economically viable. The RSM demonstrated to be more effective and reliable in finding the optimal conditions of the DAF process than the OFAT method. Thus, the RSM offered a better option than the OFAT, because it included both the interactional and individual factors.