(530c) Optimization-Based Membrane Design for Selective Nutrient Recovery from Wastewater

The release of ammonium and potassium from wastewater treatment plants is a major driver of nutrient pollution, resulting in eutrophication and harmful algal blooms in freshwater ecosystems [1, 2]. To address this issue, innovative approaches are required to recover these valuable nutrients from wastewater for fertilizer use while simultaneously minimizing environmental impact. Recent work has demonstrated that polyelectrolyte modified membranes facilitate selective recovery of these critical nutrients while rejecting the organics. Moreover, optimum tuning of the membrane properties further facilitate the partitioning of these nutrients in two separate streams - an ammonium rich permeate stream and a potassium rich retentate stream. This, in principle, enables recovering these nutrients in tailored proportions which could be used for a range of fertilizers with different N:P:K ratios. However, despite growing interest, significant challenges remain in modeling and optimizing membrane systems for nutrient recovery. These include: (i) the lack of detailed mechanistic models for membrane manufacturing processes, (ii) limited systems-level optimization studies, and (iii) the lack of software tools that enable comprehensive design and performance evaluation of nutrient-selective membranes.

This study introduces a novel membrane optimization framework and software prototype for the selective recovery of ammonium and potassium ions from anaerobic digestate. We first characterize the nonlinear relationships governing membrane behavior using an experimental dataset, encompassing membrane properties, process conditions, and performance metrics. The experimental investigations involved the use of nanofiltration membranes coated with polyelectrolyte solutions consisting of polyallylamine hydrochloride and polyacrylic acid, with glutaraldehyde (GA) as a cross-linking agent to modulate ammonium-potassium selectivity. Given the absence of mechanistic models for the layer-by-layer membrane fabrication process, data-driven regression models are employed to relate membrane permeability and selectivity to GA concentration and other fabrication parameters. A superstructure optimization approach is employed to design an efficient membrane separation system for nutrient recovery by considering the physical and structural configurations of membrane modules, including all possible interconnections (parallel, sequential, and hybrid) to achieve the desired separation performance. A mixed-integer nonlinear programming (MINLP) model is formulated to minimize membrane costs while ensuring compliance with mass balance, transport constraints, and purity specifications. The optimization framework determines the optimal GA concentration, membrane surface area, module interconnections, and overall system configuration for cost-effective nutrient recovery.

To enhance accessibility and usability, a software prototype is also developed using the Tkinter Python package [3]. The graphical user interface (GUI) enables users to specify feed conditions, membrane properties, and product specifications. Key input parameters include feed flow rates of ammonium and potassium ions, operating pressure across the membrane, and GA concentration (wt.%). The GUI automatically processes experimental data to fit nonlinear regression models and allows users to specify target recovery and purity levels for ammonium ions in the final permeate. The proposed approach provides a robust methodology for tuning membrane selectivity and optimizing system performance to achieve desired ammonium to potassium recovery, which can be tailored to meet the specific nutrient profiles of local agricultural or industrial wastewater sources. These findings have significant implications for sustainable wastewater management and fertilizer production, integrating scientific innovation with economic feasibility to mitigate nutrient pollution and promote circular economy practices in wastewater treatment.

Reference:

[1] Chislock, M. F., Doster, E., Zitomer, R. A., & Wilson, A. E. (2013). Eutrophication: causes, consequences, and controls in aquatic ecosystems. Nature Education Knowledge, 4(4), 10.

[2] Piash, K. P. S., Sprouse, N., Hodges, C., Lin, L. S., & Sanyal, O. (2025). Elucidating organic and nutrient transport mechanisms in polyelectrolyte modified membranes for selective nutrient recovery. Journal of Membrane Science, 715, 123438.

[3] Vedant, S., Atencio, M. R., Tian, Y., Meduri, V., & Pistikopoulos, E. N. (2021). Towards a software prototype for synthesis of operable process intensification systems. In Computer Aided Chemical Engineering (Vol. 50, pp. 767-772). Elsevier.