Modelovanje "cross-flow" mikrofiltracije suspenzija kvasca primenom koncepta neuronskih mreža i postupka odzivne površine

Abstract

Cilj ovog rada je ispitivanje mogućnosti primene koncepta neuronskih mreža i postupka odzivne površine za modelovanje cross-flow mikrofiltracije suspenzija kvasca. Drugi cilj je bio ispitivanje poboljšanja procesa primenom Kenics statičkog mešača kao promotora turbulencije. Primena statičkog mešača ispitana je i sa energetskog stanovišta, a ne samo sa aspekta povećanja fluksa permeata. Svi eksperimenti izvedeni su u uslovima recirkulacije i koncentrisanja napojne suspenzije. Dobijeni rezultati ukazuju da se poboljšanje mikrofiltracije može se ostvariti primenom statičkog mešača bez primene dodatne opreme. Tokom eksperimentalnog rada porast fluksa iznosio je između 89,32% i 258,86% u uslovima recirkulacije napojne suspenzije u zavisnosti od odabranih eksperimentalnih uslova, dok je u uslovima koncentrisanja napojne suspenzije porast fluksa imao vrednosti od 100% do 540% u istom eksperimentalnom opsegu. Koncept neuronskih mreža daje veoma dobre rezultate fitovanja posmatranih odziva. Pored primene ovog koncepta ispitana je i mogućnost procene uticaja pojedinih eksperimentalnih promenljivih na odzive primenom jednačine Garsona i metode jačine sinapsi koje povezuju neurone. Rezulati ovog ispitivanja u saglasnosti su sa regresionom analizom. Za detaljniju analizu uticaja eksperimentalnih promenljivih na posmatrane odzive primenjen je postupak odzivne površine funkcije. Prvi korak u ovom segmentu istraživanja bio je određivanje uticaja srednjeg prečnika pora membrane na proces mikrofiltracije. Najbolji rezultati dobijeni su za membranu srednjeg prečnika 200 nm, pošto kod većih prečnika pora dolazi do izraženijeg unutrašnjeg prljanja koje rezultuje manjim vrednostima fluksa permeata tokom proces mikroflitracije. Dalja istraživanja usmerena su na ispitivanje uticaja pojedinih eksperimentalnih promenljivih ali i njihovih interakcija za odabranu membranu (srednji prečnik pora 200 nm). Rezultati fitovanja eksperimentalnih podataka dobijeni za jednu membranu bolji su u poređenju sa rezultatima kada su fitovani eksperimentalni rezultati za sve tri korištene membrane. Sa energetske tačke gledišta primećeno je da je najbolje raditi u umerenom opsegu protoka napojne suspenzije. Kao kranji cilj primene postupka odzivne površine urađena je optimizacija vrednosti eksperimentalnih promenljivih, primenom postupka željene funkcije. Optimalni uslovi rada dobijeni u uslovima recirkulacije napojene suspenzije su transmembranski pritisak 0,2 bara, koncentracija napojne suspenzije 7,54 g/l i protok 108,52 l/h za maksimalne vrednosti specifične redukcije potrošnje energije. Sa sruge strane u uslovima koncentrisanja napojne suspenzije eksperimentalne promenljive imale su vrednosti transmembranski pritisak 1 bar, koncentracija napojne suspenzije 7,50 g/l i protok 176 l/h za maksimalne vrednosti specifične redukcije potrošnje energije.

The aim of this work was to investigate possibilities of applying neural network and response surface methodology for modeling crossflow microfiltration of yeast suspensions. Another aim was to investigate the improvement of process using Kenics static mixer as turbulence promoter. Experimental work was performed on 200, 450 and 800 nm tubular ceramic membranes. The use of static mixer was also examined from an energetic point of view not only its influence on permeate flux. All experiments were done in recirculation and concentration mode. The results clearly show that the improvement of cross-flow microfiltration of yeast suspensions performances can be done with static mixer without any additional equipment. In experimental work, flux increase had values between 89.32% and 258.86% for recirculation of feed suspension depending on experimental values of selected variables while in concentration mode this improvement was in range between 100% and 540% for the same range of experimental variables. Neural networks had excellent predictive capabilities for this kind of process. Besides examination of predictive capabilities of neural networks influence of each variable was examined by applying Garson equation and connection weights method. Results of this analysis were in fairly good agreement with regression analysis. For more detailed analysis of variables influence on the selected responses response surface methodology was implemented. First step was to investigate the influence of membrane pore size on the process of microfiltration. The results suggested that the best way to conduct microfiltration of yeast suspensions is by using the membrane with mean pore size of 200 nm, because bigger mean pore size can lead to more prominent internal fouling that causes smaller flux values. Further investigations of microfiltration process were done in order to investigate influences of variables as well as their interactions and it was done for the membrane with pore size of 200 nm. Results for this membrane considering regression analysis were considerably better compared with results obtained for modeling all three membranes. From the energetic point of view it was concluded that it is optimal to use moderate feed flows to achieve best results with implementation of static mixer. As the final goal of response surface methodology optimization of process variables was done by applying desirability function approach. Optimal values of process variables for recirculation of feed suspension were trasmembrane pressure 0.2 bar, concentration 7.54 g/l and feed flow 108.52 l/h for maximal values of specific energy reduction. On the other side for concentration of feed suspension these variables had values of 1 bar, 7.50 g/l and 176 l/h