ABSTRACT:As reverse osmosis (RO) desalination capacity increases worldwide, the need to reduce its specific energy consumption becomes more urgent. In addition to the incremental changes attainable with improved components such as membranes and pumps, more significant reduction of energy consumption can be achieved through time-varying RO processes including semi-batch processes such as closed-circuit reverse osmosis (CCRO) and fully-batch processes that have not yet been commercialized or modeled in detail. In this study, numerical models of the energy consumption of batch RO (BRO), CCRO, and the standard continuous RO process are detailed. Two new energy-efficient configurations of batch RO are analyzed. Batch systems use significantly less energy than continuous RO over a wide range of recovery ratios and source water salinities.Relative to continuous RO, models predict that CCRO and batch RO demonstrate up to 37 % and 64 % energy savings, respectively, for brackish water desalination at high water recovery. For batch RO and CCRO, the primary reductions in energy use stem from atmospheric pressure brine discharge and reduced streamwise variation in driving pressure. Fully-batch systems further reduce energy consumption by not mixing streams of different concentrations, which CCRO does. These results demonstrate that time-varying processes can significantly raise RO energy efficiency.

Superhydrophobic surfaces for enhanced condensation in Air Gap Membrane Distillation (AGMD) may provide significantly improved distillate production rates and increased thermal efficiency. While AGMD is one of the most thermally efficient membrane distillation desalination configurations, large transport resistances in the air gap limit distillate production rates. AGMD experiments were performed with combinations of untreated, hydrophobic, and superhydrophobic condensation surfaces. A nanostructured copper oxide coated condensing surface produced durable 164°±4° contact angles and jumping droplet condensation. Tests were also performed on the air gap spacer, in this case a small diameter support mesh, to judge the effects of superhydrophobic treatment and conductivity on distillate production for AGMD. A novel visualization technique was implemented to see through PVDF membranes and confirm air gap behavior. The experiments were compared with numerical modeling of AGMD film-wise condensation and flooded-gap MD. The results indicate that the introduction of superhydrophobic surfaces can result in improvements in distillate production in excess of 60% over standard AGMD. However, for high distillate production condensation on the superhydrophobic plate transitions from a partially wetted droplet morphology to Wenzel flooded (wetting) conditions. Mildly hydrophobic condensing surfaces were found to provide moderate improvement in distillate production. Superhydrophobic support meshes made a negligible difference in distillate production, but high conductivity support meshes showed significant increases in flux at the expense of thermal efficiency. The results outline recommended superhydrophobic condensation conditions at varied feed and cold side temperatures for substantial improvement to distillate production rate for AGMD systems in a flat plate configuration. KEYWORDS

Enhanced fouling resistance has been observed in recent variants of reverse osmosis (RO) desalination which use time-varying batch or semi-batch processes, such as closed-circuit RO (CCRO) and pulse flow RO (PFRO). However, the mechanisms of batch processes' fouling resistance are not well-understood, and models have not been developed for prediction of their fouling performance. Here, a framework for predicting reverse osmosis fouling is developed by comparing the fluid residence time in batch and continuous (conventional) reverse osmosis systems to the nucleation induction times for crystallization of sparingly soluble salts. This study considers the inorganic foulants calcium sulfate (gypsum), calcium carbonate (calcite), and silica, and the work predicts maximum recovery ratios for the treatment of typical water sources using batch reverse osmosis (BRO) and continuous reverse osmosis. The prediction method is validated through comparisons to the measured time delay for CaSO membrane scaling in a bench-scale, recirculating reverse osmosis unit. The maximum recovery ratio for each salt solution (CaCO, CaSO) is individually predicted as a function of inlet salinity, as shown in contour plots. Next, the maximum recovery ratios of batch and conventional RO are compared across several water sources, including seawater, brackish groundwater, and RO brine. Batch RO's shorter residence times, associated with cycling from low to high salinity during each batch, enable significantly higher recovery ratios and higher salinity than in continuous RO for all cases examined. Finally, representative brackish RO brine samples were analyzed to determine the maximum possible recovery with batch RO. Overall, the induction time modeling methodology provided here can be used to allow batch RO to operate at high salinity and high recovery, while controlling scaling. The results show that, in addition to its known energy efficiency improvement, batch RO has superior inorganic fouling resistance relative to conventional RO.