Photovoltaic powered reverse osmosis systems are stand-alone water purification systems with the potential to provide a secure drinking water supply to remote off-grid communities. These systems are often operated intermittently with extended shut-down periods to accommodate the variability of renewable energy. This work presents a novel design tool for photovoltaic powered reverse osmosis installations which considers membrane fouling during intermittent operation to ensure more reliable and long-lasting systems. The tool uses a genetic algorithm coupled to a simulation model composed of a physical system model and a cost model to evaluate the system cost. The physical model determines the local energy production. The water system model uses an experimentally validated membrane fouling model to determine the water production. Several case studies (variable location, water production, and reliability) were used to demonstrate the method. In conclusion, the results show that incorporating the fouling model was essential to design reliable and cost-optimal systems. The method showed for smaller daily water demand, the system configuration did not vary with location, indicating they can be mass-produced. The design framework will aid in the configuration of low-cost photovoltaic powered reverse osmosis systems that are reliable throughout the system life. (C) 2018 Published by Elsevier Ltd.