Numerical simulation and artificial neural network prediction of hydrodynamic and heat transfer in a geothermal heat exchanger to obtain the optimal diameter of tubes with the lowest entropy using water and Al2O3/water nanofluid

Abstract

Since the world is facing an energy crisis, and the emission of carbon-based fuel combustion is causing global warming, more research has been devoted to studying renewable energies. Due to the broad use of geothermal heat exchangers in central air-conditioning systems and also as a direct source of energy, in the present study, we have conducted a simulation of a geothermal heat exchanger and investigated the optimal diameter and nanoparticle concentration in order to minimize entropy generation. The fluid flow is turbulent, and SST K−ε is utilized as the turbulence model. The results show that as the thermal resistance of the inner wall increases, the outlet temperature rises. The effects of thermal resistance and diameter are thoroughly investigated, and the results have been used for selecting the least entropy generation. The results showed that the diameter in which the entropy is minimized increases by increasing the inner wall's thermal resistance. The best case is high thermal resistance in the outer walls and meager resistance in the inner walls. Using inner walls with high thermal resistance can decrease the adverse effects of entropy generation in large diameters. Also, the addition of nanoparticles is investigated, and the average Nusselt number and entropy generation, respectively, increased and decreased by almost 10%. Artificial neural network models are proposed to predict Nusselt number and entropy generation based on numerical results. The models are able to achieve MAE of lower than 3% and R2 higher than 0.95.