Cost-optimal design of a household batch electrodialysis desalination device

Abstract

This thesis investigates the pareto-optimal design of a household point-of-use batch electrodialysis (ED) system to provide a cost-effective replacement for existing reverse osmosis (RO) devices, for brackish water desalination of Indian groundwater, at lower energy consumption and higher recovery: 80-90% vs 25-40%. Target specifications derived from user-interviews, and RO products, guided the selection of a batch architecture, for which a coupled flow-mass transport model to predict desalination rate was developed, and validated using a lab-scale ED stack. The effects of varying the production rate (9-15 L/hr) and product concentration (100-300 mg/L) requirements on optimal selection of geometry, flow-rates, and applied voltage for total cost minimization were then explored using a multi-objective genetic algorithm. Given the low utilization of the system and the current cost of materials, the energetic cost was dominated by the capital-cost of the system. At a fixed feedwater concentration of 2000 mg/L, which is representative of the upper bound on groundwater salinity underlying much of India, and a recovery ratio of 90%, the capital cost sharply increased for systems targeted at 100 mg/L vs 200 mg/L and 300 mg/L: $141, $93, and $79, respectively averaged for systems that produced between 11.5 and 12.5 L/hr of desalinated water. Promising directions for additional cost reduction include voltage-regulation during the batch process and the development of inexpensive pumps. In addition, a candidate cost-optimal design was prototyped and tested to verify that the measured desalination performance agreed with the modeled expectations.