A generalized fractional model for one-step identification of membrane filtration clogging mechanisms

This study addresses the persistent issue of membrane fouling in filtration systems, a phenomenon that disrupts flow dynamics and reduces efficiency across various membrane types. To overcome the limitations of traditional models, a novel generalized framework, characterized by its fractional order (α), clogging indicator (n), and clogging rate constant (k), is proposed as a flexible and unified alternative to traditional models. This fractional formulation inherently allows the model to generalize effectively across various fouling behaviors: cake filtration, intermediate clogging, and standard blocking. Comparative analysis with classical models showed that the new framework consistently achieved higher accuracy, with normalized RMSE values ranging from 0.93% to 1.73% and R2 values exceeding 0.995. Due to its fractional formulation, the model demonstrates strong generalization across various fouling behaviors, without requiring separate calibration for each scenario. It also enables one-step identification and characterization of the prevailing fouling mechanism while maintaining computational simplicity. Overall, this study introduces a scalable, accurate, and robust modeling framework that enhances membrane performance in fluid and mechanical engineering applications.