P-graph model for reverse osmosis network for water processing

• Main Authors: Aguilar, Allysa Rae P., Bermejo, Jahnea Denise B., Regala, Cee Jay Z., Zenarosa, Ghielen C.
• Format: text
• Language: English
• Published: Animo Repository 2022
• Subjects: Saline water conversion—Reverse osmosis process; Chemical Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etdb_chemeng/18; https://animorepository.dlsu.edu.ph/cgi/viewcontent.cgi?article=1007&context=etdb_chemeng
• Institution: De La Salle University

Seawater desalination through the reverse osmosis (RO) technology provides cost-effective solutions for the imbalance between demand and supply of clean water. Optimized RO network designs are usually generated by a combination of heuristics approach and mathematical programming. This arrangement, however, can still be susceptible to systematic errors where other feasible and practical networks are not detected, posing a fundamental limitation. This paper introduced the P-Graph approach which demonstrated a significant contribution in realizing optimal networks and non-intuitive solutions from subjected variables. In this work, P-Graph was utilized in network design for three case studies having varied numbers of pumps, turbines, and RO modules while elements such as electricity input and flow rates were considered. Intermediate materials, such as permeate and retentate, were also evaluated in terms of design efficiency that minimized both total annualized cost and energy consumption to obtain optimal solutions. The total costs generated are $146,355, $128,128, and $535,926 with energy consumptions of 14,993.21 MJ, 9,457.78 MJ, and 30,461.51 MJ for cases #1, #2, and #3, respectively. A direct relation between the total cost and energy consumption is confirmed by the results. Thus, the P-Graph approach has the potential to optimize RO networks by recognizing all feasible network paths, including non-intuitive solutions, through different algorithms and generating total costs and energy consumption. This approach serves as a blueprint that is beneficial to engineers and decision makers, especially in large-scale applications, where accuracy, time, and flexibility in the generation of optimized networks are crucial.