مدل‌سازی ریاضی جداسازی نفت از آب با استفاده از فیلترهای ساخته‌شده از منسوجات

Introduction:
Water scarcity and pollution have emerged as critical global issues, driven by population growth, resource overexploitation, climate change, and escalating environmental threats. Among various water pollutants, crude oil and its derivatives pose significant hazards to both ecosystems and human health. Accidental oil spills during extraction, transportation, or refining processes can result in long-lasting damage to aquatic environments and severely impact biodiversity. To mitigate such risks, numerous oil removal methods have been developed, with sorbent-based approaches gaining attention for their simplicity and efficiency. Sorbent materials can selectively absorb oil from oil–water mixtures through surface and capillary interactions. Among these, technical textiles—particularly nonwoven fabrics made from polyester (PET), polypropylene (PP), and bi-component (BC) blends—have shown great potential due to their high porosity, low density, scalability, and mechanical durability. While prior research has explored their structural and surface characteristics, their dynamic behavior under varying operational conditions remains under-investigated. Mathematical modeling provides a valuable tool to simulate and predict filtration performance over time. By modeling the oil absorption process, key factors such as initial oil concentration, textile type, and flow dynamics can be quantitatively analyzed. This study designs a simple textile-based filtration system using PET, PP, and BC fabrics, aiming to model oil separation performance and evaluate each material’s behavior under different concentration levels.
Methods:
A custom-made filtration system was developed to evaluate the oil absorption performance of three nonwoven technical textiles: polyester (PET), polypropylene (PP), and bi-component (BC) fabrics. The filter body was constructed using 3-mm-thick polyethylene-based plastic sheets known for their high chemical resistance. The filter had a rectangular shape (10 × 10 × 20 cm), with a 3-cm-thick textile layer placed at the bottom to serve as the sorbent medium. Three oil–water mixtures with crude oil concentrations of 10%, 20%, and 30% by weight were prepared using tap water. Each mixture was passed through the filter, and the oil concentration in the effluent was measured at regular time intervals to monitor separation efficiency. To analyze the oil absorption behavior, a mathematical model based on the mass balance principle was developed, assuming first-order kinetics for oil sorption. The governing equation was calibrated using experimental data, and key parameters, including the rate coefficient (λ) and pollutant load (W), were estimated via curve-fitting. Model accuracy was evaluated using the coefficient of determination (R²) and normalized root mean square error (NRMSE). The validated model was then employed to predict effluent oil concentrations under various conditions, providing valuable insights into the filtration performance of the different textiles.
Results:
The proposed mass balance model effectively simulated the release of oil from textiles at different initial concentrations (10%, 20%, and 30%) for BC, PET, and PP fabrics. The pollutant load (W) and the transfer coefficient (λ) were used to evaluate the dynamics of oil transfer. The results showed that oil release varied significantly based on both textile type and concentration. BC fabrics released the highest amount of oil at all concentrations, indicating a greater tendency for oil desorption, while PP exhibited the lowest release. PET showed intermediate behavior. The model accurately estimated the pollutant load for all textile types, with R² values between 0.84 and 0.98, indicating a strong fit between observed and predicted values. The NRMSE ranged from 2.1% to 12.9%, confirming high model accuracy. Interestingly, the transfer coefficient (λ) decreased with increasing oil concentration for BC and PET, suggesting a saturation or resistance effect at higher loads. In contrast, PP showed an increasing λ trend, potentially due to differences in surface properties or oil interaction mechanisms. These patterns highlight the model’s capability to capture the distinct release behaviors of each fabric type under varying conditions. Overall, the study demonstrates the reliability and applicability of the proposed model in predicting oil release from textiles, offering valuable insight for environmental pollution modeling and textile waste management strategies.
Conclusion:
The results of this study demonstrated that technical textiles can serve as an effective and economical method for removing oil from aqueous mixtures. Among the three fabric types examined, the BC composite fabric, due to its high porosity and suitable structural composition, exhibited the highest absorption efficiency and model accuracy under most conditions. The developed mathematical model, based on the mass balance principle and assuming first-order reaction kinetics, successfully captured the dynamic behavior of oil absorption, particularly at low and medium oil concentrations. However, at higher contamination levels and with specific fabrics such as PP, the model’s accuracy decreased, indicating the need for further refinement and enhancement of the model under these conditions. Statistical analyses and comparative plots confirmed that kinetic modeling is a reliable tool for predicting the performance of oil-absorbing filters. Moreover, the findings can inform the design and optimization of industrial filters for treating water contaminated with oil and other organic pollutants. Ultimately, this study lays the groundwork for developing more advanced modeling approaches and the application of environmentally friendly technologies in managing oil pollution.