Publication

Abstract : The rapid growth of industrialization and the depletion of global water resources have made the treatment and reuse of hazardous industrial effluents crucial for sustainable water management, directly supporting the Clean Water and Sanitation Sustainable Goals. As conventional treatment technologies have continued to be both expensive and energy consuming, biologically assisted technologies present environmentally friendly and scalable alternatives. This research has developed a new bio-rheological modeling framework to simulate the flow of effluents and the mechanisms used to transport pollutants based on microrotation induced by microbial motility combined with the non-Newtonian Casson fluid behavior of effluents. The bio-rheological model incorporates the complex physical, thermal, and biologically based interactions of microbial motility, heat, and hazardous contaminant transport through molecular-based nonlinear radiation, Arrhenius rate law reaction kinetics, and thermo-diffusive effects. Using a Central Composite Design based deep neural network (DNN) model, we can predict key engineering quantities such as heat and mass transfer rates, and rates of microbial motility with greater accuracy and efficiency than achieved with conventional analytical and empirical methods. In fact, the DNN model produced R² values of 0.9989 for the heat transfer rate, 0.9996 for the mass transfer rate and 0.9993 for the density of microbial motility, which demonstrates that the DNN provided an accurate representation of the real-world physical, thermal, and microbiological interactions between effluents and their associated pollutants. This integrated biofluid-AI approach provides practitioners with a data-driven and sustainable framework to optimally design and implement biological treatment systems for industrial effluent with minimal cost, energy consumption, and environmental impact.