The relationship between water and energy resources is complex and interdependent, commonly referred to as the water-energy nexus (WEN) [1]. The demand for freshwater and energy is increasing, which has led to concerns about the United Nations Sustainable Development Goals (SDG). The SDG 6 (Clean water and Sanitation) and the SDG 7 (Affordable and Clean Energy) are particularly threatened [2], [3]. Global population growth will have a significant impact on energy and water demand and will place significant strain on both resources[2].To address these challenges, integrated and smart management of water and energy resources is crucial at every scale, from the national to the industrial plant scale. Water reuse or recycling system with low environmental impact has transitioned from being a choice to a necessity. In this context, TotalEnergies has deployed a pilot-scale water treatment platform (SWAP: Sustainable Water Platform), a dual-membrane-based ultrafiltration and reverse osmosis water treatment powered by PV-T panels with integrated co-use of heat and electricity. Such platform aims to assess the synergy potential between energy generation and water treatment processes. The SWAP pilot also aims to demonstrate that all available water sources on or near TotalEnergies’ sites can be used (recycling or reuse) as an alternative to traditional water sources, with low CO₂ direct emissions and environmental impacts. The present work aims to develop a digital twin of SWAP before considering operational optimization.

For membranes performance and fouling modeling purpose, Resistance-in-series model[4] is chosen to represent the hollow fiber Ultrafiltration skid dynamics facing different operating conditions such as various flux and water qualities (from treated wastewater after biological treatment to collected rainwater). Solution-diffusion model[5] coupled with the film theory is assumed for reverse osmosis skid modeling. Models’ predictions are compared to SWAP pilot experimental data for validation.

To operate a water treatment plant powered by renewable energy and make the most out of it, it is pertinent to rely on an optimization problem formulation[6]. Optimization techniques have been widely applied in multi-energy systems based on the water-energy-nexus[7], [8], [9]. Yet, in water treatment, the focus has been mainly on optimizing desalination, with limited integration of environmental dimension. The present study explores also SWAP dynamic offline operational optimization. A special focus over the low-grade fatal heat revalorization by filtration skids is assessed. The medium-term dynamic optimization aims to maximize treated water production while minimizing operational cost and environmental impacts. Optimization parameters are the raw water quality, on-site water needs and weather log data. The optimization variables are the skids feeding flux. Expected optimization results are the system optimal operating trajectories over 3-5 days’ time horizon. This typically includes ...