Studying evaporation from Ottawa sand with mixed wettabilities under a simulated solar flux with forced convection: experimental apparatus design and evaporation rates

Abstract

The primary source of water for crops and livestock in the United States Central High Plains is irrigation from the Ogallala Aquifer [1], supplying 30% of the water used in the United States’ irrigated agriculture in 2015 [2]. Due to the semi-arid climate of this region, limited rainfall (e.g., 33-74 cm of precipitation [3]) contributes to watering crops, thereby resulting in water scarcity. Reducing the evaporation from soil is one approach to conserving water and improving water use efficiency. In this study, a soil-water evaporation test section was designed and constructed to study evaporation from Ottawa sand with constant air flow above and below the sand layer. A simulated solar flux was applied to the sand, while varying soil wettability (e.g., all hydrophilic, 12% hydrophobic sand concentrated in a layer, and 12% hydrophobic sand evenly mixed). Prior to entering the sand test section, compressed air was dried in a desiccator and then split in two flows before flowing through the 5.7-cm-depth, 22.8-cm-wide, and 83.8-cm-long test section, with one air stream flowing above the 5.7-cm-thick sand layer and the other below and, subsequently, flowing through the moist sand layer. The air temperature, pressure, and relative humidity were measured at the inlets and outlet of the test section to measure the change in moisture and, therefore, water content removed from the sand via evaporation. Using inlet air mass flow rates of air of approximately 2E-4 kg/s for each inlet, air temperatures of 28–31 °C, and dry air (i.e. 0–2% RH), an applied solar heat flux of 112±20 W/m², and steady state exit air temperatures of approximately 27 °C, evaporation rates were measured for three cases: 1) a 5.7-cm thick layer of hydrophilic Ottawa sand; 2) a 5.7-cm-thick layer with 12% hydrophobic content consisting of a 0.7-cm-layer of hydrophobic sand buried 1.8 cm below the surface of hydrophilic sand, termed hydrophobic layer; and 3) a 5.7-cm-thick layer with mixed wettabilities consisting of 12% hydrophobic sand mixed into hydrophilic sand, termed hydrophobic mixture. Evaporation from porous media often occurs at a high, constant-rate driven by capillary flow for higher saturations. This is followed by a falling-rate, and slow-rate evaporation driven by vapor diffusion for lower saturations. The saturation percentage of the sand was measured via a gravimetric approach during evaporation of the three experiments. The evaporation rate of water for each experiment began at a quasi-constant rate: approximately 8.7E-6 kg/s to 9.6E-6 kg/s for the all hydrophilic case, 8.6 E-6 kg/s to 8.9 E-6 kg/s for the hydrophobic layer, and 8.1E-6 kg/s to 9.0E-6 kg/s for the hydrophobic mixture. The hydrophilic experiment entered the falling-rate of evaporation at 12% saturation, the hydrophobic layer entered the falling-rate of evaporation at 20% saturation, and the hydrophobic mixture entered the falling-rate of evaporation at 24% saturation. The change in onset of the falling-rate of evaporation led to more experimental measurements required to complete the entire saturation curve for the hydrophobic experiments. With each experiment beginning at approximately 40% saturation, the hydrophilic experiment took 17 trials to complete, the hydrophobic layer took 20 trials to complete, and the hydrophobic mixture took 26 trials to complete. The hydrophobic mixture took the most trials to complete the saturation curve and had the shortest constant-rate period of evaporation. The addition of hydrophobic Ottawa sand to hydrophilic Ottawa sand reduced the evaporation rates at higher saturation percentages, thereby increasing water retention compared to all hydrophilic Ottawa sand. Effective thermal conductivities were calculated for each experiment at three locations within the sand layer. Effective thermal conductivities were found to be higher than the expected values. This was attributed to uneven moisture distribution and thermal losses by convection and conduction through the test section walls. The evaporation flux was measured to be up to 12 times higher than the diffusion flux calculated throughout each experiment. This suggests that capillary action, thermal gradients, and convection are primary drivers of evaporation and not vapor diffusion.