MINLP Moel for Reverse Osmosis Network Design uner Time-Variant Operation Constraints

Reverse osmosis (RO) is a very well-establishe technology for water esalination. Commercial RO systems are compose of several numbers of RO passes with auxiliary equipment such as high-pressure pumps an pressure-exchanger units to facilitate the separation, an to eliver clean water. Available esign moels for the RO network represent mixe integer nonlinear programming (MINLP) formulations. In aition, the existing RO esign MINLP moels are static in nature. These moels are not suitable for RO network esign uner time-variant constraints. In this stuy, a multiperio MINLP moel is etaile for the RO network esign. High-level esign ecision variables are relate to the RO network configuration an equipment sizing, which are inepenent of time. These ecisions are moele by iscrete an continuous variables. Low-level time-variant ecision variables are relate to the RO network operation an moele by continuous variables. The operation ecision variables are constraine by the esign variables through ifferent time perios to satisfy water-eman constraints. In aition, several correlations are consiere in this stuy to estimate water an RO transport properties uring the separation process. Case stuies are analyze to show the application of the mathematical programming moel.