RO Process Optimization Based on Deterministic Process Model Coupled with Stochastic Cost Model

Abstract

A survey performed over existing two pilot-scale and two full-scale RO desalination facilities to study the current status of boron rejection showed a highest rejection 85% leading to permeate boron concentration of 0.52 mg/L, and recent studies predicted a cost increase due to incorporation of boron reduction systems. Mathematical models were developed to study the process performance and related cost implications. The deterministic process model was verified with pilot-scale experiment performed using a single spiral wound module and was later modified to represent the full-scale design options available to meet the required water quality criteria. Then the selected full-scale design options were simulated to predict their performance in terms of recovery and boron rejection. For cost analysis, to account for uncertainty probability models were developed for stochastic inputs to the cost estimation model and were used with operating parameters from the full-scale simulations to determine the expected total cost of water produced. Later, a sensitivity analysis was performed to observe the effect of change in uncertainty of inputs. Further, the applications of the deterministic process model are suggested.