Anaerobic membrane bioreactors for domestic wastewater treatment: treatment performance and fouling characterization

Abstract

As the human population continues to grow, increasing stress is placed on the food-energy-water nexus. Wastewater is increasingly being viewed as a sustainable resource to close the loop in this nexus as it contains nutrients, latent chemical energy, and water that can be recovered for beneficial reuse. One technology that has shown promise in harnessing wastewater as a resource is the Anaerobic Membrane Bioreactor (AnMBR). AnMBRs combine anaerobic biological wastewater treatment with membrane filtration technology to achieve superior effluent quality, while generating energy in the form of methane and providing an opportunity to capture nutrients. In this study, a pilot scale gas-sparged AnMBR was operated in Ft. Riley for 472 days under ambient temperature conditions. The AnMBR achieved an average removal efficiency of 88±7% and 88±6% for COD and BOD₅, respectively, at temperatures ranging from 12.7°C to 31.5°C, demonstrating its feasibility for ambient temperature operation. The AnMBR was also paired with downstream nutrient recovery using a coagulation-flocculation-sedimentation process, which removed 94±3% of phosphorus and over 99% of nitrogen, as well as both gaseous and dissolved methane capture, which could generate an estimated 72.8% of the power required for energy neutrality. The successful integration of AnMBRs in a treatment train that addressed the critical challenges of dissolved methane and nutrient capture demonstrates the viability of the technology in achieving holistic wastewater treatment. While the viability of the pilot scale AnMBR for municipal wastewater resource recovery was successfully demonstrated, membrane fouling was identified to be a major obstacle to the widespread adoption of the technology. In the pilot system, fouling management through gas sparging required 54% of the total energy demand, which could be reduced with improved targeted maintenance strategies that can be developed after a deeper understanding of the foulants. To better understand fouling behavior in AnMBRs, a suite of analyses was employed to dynamically monitor the initiation and proliferation of fouling to complement the traditional end-of-life membrane autopsies in the pilot-scale AnMBR. Fluorometry and soluble chemical oxygen demand (sCOD) monitoring of the membrane permeate during steady state pilot AnMBR operation complemented end-of-life membrane and sludge cake analyses, such as Fourier Transform Infrared Spectroscopy (FTIR), to identify both organic and inorganic foulants that have been deposited on a membrane surface throughout the pilot operation. The fouling events coincided with spikes in the permeate sCOD, accompanied with an increase in tryptophan and tyrosine like compounds, as measured by the fluorometric excitation-emission (EEM) spectra. Scaling, albeit minor, was mainly accounted by calcium sulfate and calcium phosphate, as opposed to the typically expected calcium carbonate. These findings have implications for optimizing membrane maintenance strategies, informing solids wasting schedules as well as the selection of chemicals for backpulsing. To further mechanistically understand membrane fouling, a novel lab scale AnMBR with a side tube membrane module was designed to characterize early membrane foulants. A main module was continuously operated to maintain system performance, averaging 83±7% COD removal during the startup period, while the side tube module was installed and operated in parallel to allow for sampling the membrane fibers without interrupting the main module. The lab scale system underwent extended critical flux testing while being fed with synthetic municipal wastewater only, followed by controlled spikes of tryptophan, tyrosine, and humic acid into the wastewater. The transmembrane pressure hysteresis was much more severe in the spiked test. Endpoint analysis after sub-critical flux operation revealed that cake formation was limited in the side tube samples suggesting that the hysteresis and initial fouling was due to pore constriction, which is a significantly less studied fouling mechanism compared to cake layer fouling. Cake layer formation in the main module was localized in the top third of the membrane fibers. Heat extraction temperature profiles that were optimized in a separate study were applied for soluble microbial product and extracellular polymeric substance characterization of the membrane extracts and cake foulants after each critical flux test. Humic and fulvic acids appear to be more associated with pore fouling, while proteins appear to be more associated with the cake layer. Free amino acids were not detected in any sample suggesting that the amino acids likely underwent rapid anaerobic transformation. Tryptophan-like and tyrosine-like fluorescence was found in the EEMs of the side tube module extracts and were likely due to amino acid degradation products, such as indoles, due to having peaks in the EEM but the spiked test not increasing in protein concentration. In contrast, the foulant cake’s proteinaceous foulants were likely in the form of polypeptides, as confirmed through FTIR and colorimetry. Significant amounts of humic substances in the form of humic and fulvic acids were extracted from the membrane surface and pores and likely also contributed pore fouling, with significantly higher amounts being extracted from the side tube membrane module after the spiked test and the main membrane module at the end of operation. These findings show great promise for guiding future research on early onset of membrane fouling and fouling management strategies to achieve energy positive operation using the AnMBR technology platform.