Season-Wise Three-Dimensional CFD Modelling of a Natural Evaporative Water Cooler: Development, Calibration and Experimental Validation

Abstract

Natural evaporative water coolers provide refrigerant-free, ultra-low-energy water cooling for hot–dry climates, yet computational fluid dynamics (CFD) studies of such systems at full device scale are absent from the literature. This paper develops a three-dimensional CFD model of a complete natural evaporative water cooler in OpenFOAM and validates it season-wise against the authors' experimental campaign at the Taguchi-optimised operating point of each season. The water side resolves the entire 76.2 m helical copper coil (inner diameter 4.5 mm, 80.85 turns) with a structured butterfly O-grid mesh of 155,200 hexahedral cells generated by a purpose-built sweep algorithm; the wetted evaporative pad is represented by an external film boundary condition referenced to the ambient wet-bulb temperature. A companion air-side model treats the pad–coil assembly as a Darcy–Forchheimer porous zone inside the cooler housing. The film coefficient was calibrated once, at the summer optimum only (h = 41 W/m²·K); the same value was reused unchanged for the winter simulation — making winter a genuine blind prediction — while the monsoon value (h = 28 W/m²·K) was estimated a priori from the resistance network for the reduced sprinkle rate and humid air. The model reproduces the measured seasonal temperature drops of 9.7, 5.4 and 5.6 °C with deviations of −0.5%, −1.6% and −1.2% for summer, monsoon and winter respectively, and cooling effectiveness within 2 percentage points of experiment in all seasons. The calibrated wetted-pad conductance is 1.9 times the dry-pad analytical estimate, quantifying the contribution of direct film evaporation, and the resolved Dean vortices (De ≈ 340) explain the weak experimental sensitivity to coil water velocity in summer.