(364a) Energy-Optimal Control of An Integrated UF-RO Seawater Desalination Plant

In recent years, reverse osmosis (RO) desalination has emerged as a leading method for desalting seawater, inland brackish water, and water reuse applications due to the versatility of this process relative to other water desalination technologies.  Large-scale RO desalination plants are common in many parts of the world today.  Despite this, as popularity of the process grows, there is a growing demand for small-scale distributed systems.  Minimizing energy use is always a concern when optimizing RO plant operation but energy recovery devices are impractical for small-scale systems low efficiencies.

In order to carry out energy-optimal operation of a small-scale RO desalination plant, it is important that the control system be able to maintain freshwater production and energy-optimal operating conditions simultaneously.  In order achieve all this, a novel model-based RO controller was developed at UCLA.  This controller was then implemented into a novel RO system integrated with its ultrafiltration (UF) pre-treatment system also built at UCLA and field tested at Port Hueneme US Naval Base. 

Previous research has been done by Zhu et al. [1] that demonstrate the use of the specific energy consumption (SEC) with respect to RO water recovery as a framework to determine energy-minimal RO operating conditions.  Further analysis is done in this framework in order to develop a simple algorithm to calculate the energy-minimal operating point.  The analysis shows that the energy-optimal operating condition always occurs at the maximum physically possible RO water recovery. 

This algorithm is combined with a simple RO model which is derived from the classical RO flux equation.  This model calculates the RO feed flow rate and the RO feed pressure required to achieve the RO water recovery and RO permeate flow rate set-points.  These set-points are then communicated to local controllers for actuation. 

Experiments were carried out to observe the performance of this novel model-based controller.  It was demonstrated that the controller was successful in accurately achieving the RO permeate flow rate and RO feed pressure set-points.   It was also shown that the controller was able to independently control the two set-points even though the control variables are dynamically coupled.  Disturbances were induced through unsteady-state operation of the UF pre-treatment system and it was demonstrated that the system successfully rejected the disturbance by maintaining its set-points despite fluctuations in the feed water conditions.  It was concluded after the study that this novel model-based controller enabled robust control of an integrated UF-RO system under energy-optimal conditions. 

1. Zhu, A., P. D. Christofides and Y. Cohen, “Effect of Thermodynamic Restriction on Energy Cost Optimization of RO Membrane Water Desalination,” Industrial and Engineering Chemistry Research, 48, 6010-6021 (2009).