Autonomous control and membrane maintenance optimization of photovoltaic reverse osmosis systems

Abstract

The supply of clean water in remote and off-grid areas has been a major global challenge for humanity. Over 780 million people lack access to clean water [1]. However, a significant fraction of these people have access to undrinkable surface, brackish or sea water. A promising solution to this problem is to use photovoltaic powered reverse osmosis (PVRO) systems to purify this unsafe water to produce clean drinking water. However, high initial capital costs and a lack of commercial viability have prohibited these systems for commercial and daily use. For this approach to be feasible and reach large-scale commercial viability, PVRO systems need to be energy efficient and cost-competitive compared with reverse osmosis systems powered by conventional sources, such as diesel engines or electricity from the grid. The costs and energy consumption in a PVRO system can be significantly decreased by maximizing water production and minimizing the effects of membrane degradation to extend system life. The membrane degradation considered here is the fouling phenomenon in which suspended solids and dissolved substances collect on the surface and within the pores of the membrane thereby reducing its permeability This thesis describes an innovative approach to autonomously controlling and optimizing community scale PVRO systems by controlling membrane degradation due to fouling, using a self-optimizing condition based maintenance algorithm. Additionally, by exploiting the energy compliance of PVRO elements and actively controlling the individual components of the system, water production can be maximized. The compliance in a PVRO system has been found to significantly affect PVRO performance by reducing system efficiency and resulting in long startup delays in producing clean water. In this thesis, a controllable recovery ratio concept system has been presented. By actively controlling the PVRO system, an improvement of 47% over the existing performance of a fixed recovery ratio system has been shown in simulations. Use of condition based maintenance strategies show an improvement of over 10% in cumulative clean water production compared to scheduled quarterly maintenance and 58% over 1 year in cumulative clean water production compared to the case without any maintenance. This is interesting since typical community scale and point of use systems can be and are operated without periodic maintenance [2]. Combining the optimal power control and condition-based maintenance strategies, an improvement in water production of 85 % is shown for a July day in Boston over the MIT PVRO system. Finally, a self-optimizing condition based maintenance algorithm is proposed as the optimal solution to control membrane degradation due to fouling.