(658f) Determination of membrane pore size distribution using nonlinear regression

Pore size and pore size distribution (PSD) are significant parameters affecting membrane performance and efficiency. In filtration processes, larger pore radii result in higher flux and lower rejection. In Membrane Distillation (MD), larger pore radii can result in higher flux and penetration of liquid into the pore and occurrence of pore wetting. There is a trade-off between the occurrence of pore wetting and high flux. Therefore, an optimum pore radius should be selected depending on different parameters such as feed type, temperature, etc. The critical point is that for MD membranes, the radius of pores should be distributed in a narrow region around the optimum radius (narrow PSD) because even a small number of large pores may cause pore wetting and consequently reduce flux and/or permeate quality. Hence, estimation of MD membrane PSD is of great importance. Among various available methods, the bubble gas transport method is a common approach for PSD estimation. However, this technique requires many experimental data points. With the proposed model, PSD can be predicted with only four data points, reducing the effort of recording unnecessary data. Unlike the conventional approach, the proposed approach starts from assuming a distribution function for the PSD, according to which a flow rate versus pressure curve is produced. Then, the mean pore size and standard deviation of the mentioned function are optimized by the best fit of the theoretical curve to the experimental data through minimizing an error function using nonlinear regression.