| dc.description.abstract | In countries such as India, where continuous access to treated piped-water is uncommon, many have resorted to desalinating brackish groundwater to meet their drinking needs. This form of decentralized treatment is performed at the community-scale, as is common in rural areas, and within individual homes, using point-of-use (POU) purifiers. This thesis develops methods to lower the costs and improve the efficiencies of two technologies for these applications: electrodialysis (ED) and reverse osmosis (RO).
Batch ED desalination, which relies on recirculating water to reach a desired product concentration, is often conducted at constant voltage. This operation scheme causes the membrane area to be underutilized because the ratio of applied current to limiting current is initially low during the batch cycle. By applying a time-varying voltage to the ED stack, we raised this ratio and increased production rate by up to 37% using the same membrane area. In parallel, we derived an analytical prediction of the batch time and validated it under varying feed and product concentrations, and flow velocities. The experiments and model together suggest that the proposed control scheme will improve production rate most significantly when desalinating through large concentration changes at low flow velocities. This work will assist engineers and operators seeking to size, evaluate, and maximize the production performance of new and existing batch ED systems.
Decreasing the energy requirements of community-scale RO, by recovering hydraulic power from the brine stream, will make off-grid deployments more affordable. However, existing energy recovery devices (ERDs) are prohibitively expensive. We investigated the feasibility of leveraging ubiquitous gear and sliding vane positive-displacement mechanisms within a fixed-recovery architecture to provide a low-cost ERD solution. By modeling the coupled behavior of the pump, ERD, and RO train, we showed that production performance is sensitive to volumetric efficiency. Based on this finding, vanes were selected over gears for prototyping. The prototype enabled a 17% decrease in measured power consumption, and through characterizing friction, we determined that these savings could be doubled by balancing pressure loads on the vane mechanism’s rotor. This work lays the groundwork for realizing an affordable ERD for community-scale RO treatment.
Finally, today’s POU RO purifiers only recover 20-30% of the input feed as drinking water and consume significant energy. By testing and analyzing a POU RO system, it was identified that recirculating the brine within a semi-batch configuration could help address these limitations. We engineered such a system using off-the-shelf parts, and in initial testing, showed that it could achieve recoveries of up to 75% without affecting production rate and quality. With further testing and refinement, this semi-batch system could make POU water desalination more efficient. | |
