Optimising Metformin Degradation via Hydrodynamic Cavitation: A Comparative Study of Venturi and Orifice Devices Using Response Surface Methodology

Abstract

Background: The Hydrodynamic Cavitation (HC) is a promising, energy-saving method to remove micropollutants such as metformin in water. There are however difficulty in optimisation of the performance of HC reactor as the cavitation behaviour is geometry dependant and the operating parameters interact in complex and non-linear ways. Nevertheless, there is still a challenge in optimising reactor performance among various device geometries (Venturi and orifice) because there are intricate interactions amongst operating parameters. Objectives: The objective of this study is to compare and evaluate systematically the degradation performance of Venturi and orifice-type HC devices to remove metformin. It uses the Response Surface Methodology (RSM) to model, optimise and interpret the influence of the essential parameters such as pH, pressure drop, treatment time and device type within experimentally justified operating ranges relevant to pharmaceutical wastewater treatment. Materials and Methods: It conducted 31 experiments according to a Central Composite Design (CCD) structure. The variable of response was the degree of degradation of metformin. Statistical modelling was done in multiple regression, ANOVA and analysis of the interaction effect. The energy yield, cavitation intensity and the cost per gram of TOC removed were also calculated and degradation kinetics were evaluated using a pseudo-first- order kinetic model. Results: The most significant factors (p<0.001) were pH and pressure drop with strong quadratic and interaction effects. The maximum rates of degradation occurred at pH ≈ 4.5 and ΔP ≈ 4.2 bar, with treatment time showing diminishing returns beyond ~50 min. Venturi devices were more efficient in degradation, cavitation yield (mg/J) and kinetic rate constants than orifices achieving approximately 1.8-2.2 times higher degradation under comparable operating conditions. The cost analysis showed that the Venturi system had a lower operating cost of about 25%. There was an insignificant effect of aeration. Conclusion: High-fidelity process optimisation was made possible on both of these devices using RSM. Although Venturi designs are more efficient, the classical HC devices are still suitable in the case of decentralised treatment because of the simple design and cost-effectiveness. The findings demonstrate practical scalability and applicability of HC-based systems for pharmaceutical wastewater treatment. Future directions need to be in hybrid geometries, actual wastewater matrices, and scale-up.