(604e) Modeling, Design and Upscaling the Membrane Bioreactor Process for Removing Biodegradable Organic Matter and Disinfection Byproduct Precursors from Water Supplies

This research discusses the use of membrane bioreactor (MBR) technology for the removal of biodegradable organic matter (BOM) and disinfection byproduct (DBP) precursors from potable water sources after ozonation. It is indeed well known that ozonation of natural organic matter (NOM) in water supplies results in the cleaving of the organic macromolecules into more easily biodegradable organic matter including aledehydes and ketones. Subsequent to advanced oxidation processes such as ozonation, the BOM essentially consists of smaller molecules that can serve as organic substrates for microbial growth, and can constitute organic molecules that can generate DBPs such as trihalometahnes and haloacetic acids. The purpose of BOM removal after ozonation using the MBR technology was twofold: preventing the growth or regrowth of pathogenic microorganisms in the water distribution system; and reducing the potential for forming DBPs such as trihalometahnes and haloacetic acids.

The removal of BOM was measured in terms of three variables, namely: dissolved organic carbon (DOC), assimilable organic carbon (AOC), and total aledehydes. These parameters were measured after ozonation, and biological treatment in MBR following ozonation. Besides these variables, the DBP formation potentials were also measured in terms of trihalomethane formation potential (THMFP). The AOC was determined using Pseudomonas flourescens P-17 and Aquaspirillum NOX bacterial strains as bioassay organisms. The total aledehydes denoted the total concentration of formaldehyde, acetaldehyde, glyoxal, and methyl glyoxal.

The MBR experiments established the effectiveness of the MBR technology in achieving excellent removals of BOM and DBP precursors. Both AOC and total aldehydes were completely removed under optimal operating conditions, while the DOC removal was 60 percent. The studies further demonstrated showed that PAC application not only enhanced DOC removal but also greatly increased the membrane permeate flux, improving the overall economic viability of the process.

A predictive model was developed for performance forecasting and design of the MBR process. A lumped parameter approach was used in using the three variables for representing BOM (DOC, AOC and total aledehydes). The model involved the phenomenological aspects pertaining to pollutant transport, sorption equilibrium, and biochemical reaction. It considered film transfer from bulk liquid phase to the biofilm-liquid interface, biofilm diffusion, and diffusion into the adsorbent particle. It also considered biochemical reaction in the biofilm immobilized on the PAC adsorbent particle, and in the aqueous suspended phase. These reactions described biofilm and biomass growth and decay within the MBR system, and were governed by Monod kinetics. The adsorption equilibrium and kinetic relationships controlled the sorption of BOM from the aqueous phase, and their release from the adsorbent to the biofilm as well as to the microorganisms in the suspended phase. The BOM removal mechanisms therefore involved a combination of biofilm degradation, suspended phase biodegradation, and adsorption.

The MBR model parameters were determined from independent laboratory-scale experiments and correlation techniques for each of the three BOM variables of interest (DOC, AOC and total aledehydes). The adsorption equilibrium and rate parameters were obtained from batch reactor studies, while the biokinetic parameters were estimated from chemostat studies. The MBR experimental data provided model verification under different process conditions. Simulation studies provided a qualitative evaluation of the parameters influencing the MBR process dynamics under different conditions. These parameters involved liquid film mass transfer, biofilm transport, biological degradation kinetics, influent concentration, and reactor hydraulic retention time.

A new feature presented in the modeling approach is the evaluation of MBR process dynamics under certain real scenarios including process shutdown, variability in influent concentrations of target contaminants, and changes in hydraulic retention time, as well as process upscaling. Process shutdown often upsets biological systems owing to unpredictability in the behavior of microorganisms. Variations in influent concentrations of contaminants arise from fluctuations in the NOM or BOM content of the treated water due to changes in water quality. Changes in hydraulic times are attributed to permeate flux decline arising in membrane systems operated under constant trans-membrane pressures owing to membrane fouling and concentration polarization effects. Besides these aspects, process upscaling by dimensional analysis and similitude are also investigated.

Key words: membrane bioreactor, modeling, dissolved organic carbon (DOC), assimilable organic carbon (AOC), total aledehydes, disinfection byproducts (DBPs), biodegradable organic matter (BOM), process variables and dynamics, upscaling