Renewable energy systems coupled with reverse osmosis (RO) desalination represent a great contribution to the fields of energy and sustainability. Especially in remote areas, where reliability issues associated with integrated systems result in problems of insufficient power supply or supply loss due to the intermittency of renewable sources and varying freshwater demand. This research aims to design a sustainable and reliable hybrid renewable energy system (HRES) coupled with RO desalination system (HRES-RO), considering different operational scenarios with fluctuating renewable energy supply, and changeable water demand. First, future energy supply from renewable sources, and water demand were forecasted to deal with the stochastic behavior of several variables including freshwater demand, ambient temperature, solar radiation, and wind speed by means of recurrent neural networks (RNN). Then, multi-criteria optimization was conducted using extended mathematical programming (EMP) with the aim of minimizing the total annual costs and greenhouse gas emission. Finally, the potential loss of power supply probability (PLPSP) was introduced as a tool to illustrate the sustainability of the proposed scenarios. The results showed that the proposed framework resulted in an HRES with optimized installation strategy using 111 photovoltaic panels and 5 wind turbines by considering three criteria (economic cost, environmental effect, and energy reliability). The designed system reduced PLPSP by 18.3% compared with the base case. Furthermore, the results demonstrated the contribution of advanced forecasting algorithms to address future uncertainties in the energy supply chain.
- Energy reliability index
- Hybrid renewable energy system
- Multi-objective optimization
- Recurrent neural network
- Supply chain forecasting model
Li, Q., Loy-Benitez, J., Nam, K. J., Hwangbo, S., Rashidi, J., & Yoo, C. (2019). Sustainable and reliable design of reverse osmosis desalination with hybrid renewable energy systems through supply chain forecasting using recurrent neural networks. Energy, 178, 277-292. https://doi.org/10.1016/j.energy.2019.04.114