In this paper, nonlinear system identification of the activated sludge process in an industrial wastewater treatment plant was completed using adaptive neuro-fuzzy inference system (ANFIS) and generalized linear model (GLM) regression. Predictive models of the effluent chemical and 5-day biochemical oxygen demands were developed from measured past inputs and outputs. From a set of candidates, least absolute shrinkage and selection operator (LASSO), and a fuzzy brute-force search were utilized in selecting the best combination of regressors for the GLMs and ANFIS models respectively. Root mean square error (RMSE) and Pearson's correlation coefficient (R-value) served as metrics in assessing the predicting performance of the models. Contrasted with the GLM predictions, the obtained modeling results show that the ANFIS models provide better predictions of the studied effluent variables. The results of the empirical search for the dominant regressors indicate the models have an enormous potential in the estimation of the time lag before a desired effluent quality can be realized, and preempting process disturbances. Hence, the models can be used in developing a software tool that will facilitate the effective management of the treatment operation.

Colloidal clogging is typically studied in pores with constrictions arranged in parallel or series. In these systems, clogging statistics are governed by Poisson processes; the time interval between clogging events exhibits an exponential distribution. However, an entirely different phenomenon is observed in a gently tapered pore geometry. Unlike in non-tapered constrictions, rigid particles clogging tapered microchannels form discrete and discontinuous clogs. In a parallelized system of tapered microchannels, we analyze distributions of clog dimensions for different flow conditions. Clog width distributions reveal a lognormal process, arising from concurrent clogging across independent parallel microchannels. Clog lengths, however, which are analogous to growth time, are exponentially distributed. This indicates a Poisson process where events do not occur simultaneously. These two processes are contradictory: clogging events are statistically dependent within each channel while clogs grow simultaneously across independent channels. The coexistence of Poisson and lognormal processes suggests a transient Markov process in which clogs occur both independently of, and dependently on, other clogs. Therefore, discussions of the stochastic character of clogging may require holistic consideration of the quantities used to assess it. This study reveals small adjustments to pore spaces can lead to qualitative differences in clogging dynamics, suggesting the importance of geometry.

This study demonstrates a method to inhibit the growth of CaCO3 and synthesize submicron particles in a chemical precipitation process, under ambient and high supersaturation conditions. Equimolar CaCl2 and Na2CO3...

Suspended particles flowing through complex porous spaces exhibit clogging mechanisms determined by factors including their size, deformability, and the geometry of the confinement. This study describes the clogging of rigid particles in a microfluidic device made up of parallel microchannels which taper from the inlet to the outlet, where the constriction width is approximately equal to the particle size. This geometry summarizes the dynamics of clogging in flow channels with constrictions that narrow over multiple length scales. Flow tests are conducted at constant driving pressures for different particle volume fractions, and a power-law decay which appears to be peculiar to the tapered geometry of the channels is observed in all cases. In comparison with non-tapered channels, the power-law exponent shows the flowrate decay rate is significantly lower in a tapered channel. Also, micrographs of the clogged channels reveal the clogs do not grow continuously from their initial points of inception. Rather, multiple clogs with increasing number of particles in their cross section are successively formed as the cake grows in each microchannel. Changes in particle volume fraction at a constant driving pressure affect the clogging rate without impacting the underlying clogging dynamics. Unexpectedly, analyses of the particles packing behavior in the microchannels and post-clogging permeability reveal the presence of two distinct regimes of driving pressure, though only a small portion of the total device volume and channels surface area were occupied by clogs, regardless of the particle volume fraction.