TOWARDS TERTIARY MICROPOLLUTANTS REMOVAL BY BIOAUGMENTED MOVING BED BIOFILM REACTORS (MBBRS) AND NANOFILTRATION (NF)
This doctoral degree program was undertaken together with the Université Paul Sabatier – Toulouse III (France) and the University of Leuven (Belgium). Therefore Seyed Mehran Abtahi Foroushani will defend his dissertation at the Université Paul Sabatier – Toulouse.
Seyed Mehran Abtahi Foroushani is a PhD student in the research group Membrane Science and Technology. His supervisor from the University of Twente is prof.dr.ir. H.D.W. Roesink from the faculty of Science and Technology.
Prevailing trends in global development, specifically the increases in population, urbanization, economic welfare, and use of chemicals, results in increased pressures on the quality of water. Today, when water supply intakes are downstream of wastewater treatment plants, most existing water quality standards are met while micropollutants (MPs) of the treated effluent are often seen as a serious problem. To reduce the release of such compounds into the surface waters, development of tertiary treatment technologies has been noticeably heeded over the last decade. To broaden such knowledge, two advanced treatments called “bioaugmented moving bed biofilm reactors (MBBRs)” and “polyelectrolyte multilayer (PEM)-based nanofiltration (NF) membranes” were studied in this thesis to elucidate their potential for the elimination of several MPs from conventionally-treated municipal wastewater.
- Tertiary MBBRs
Three identical glass-made MBBRs, each with an effective volume of 3.1 L, were continuously fed by a synthetic MPs-bearing secondary-treated wastewater, and operated in parallel under the ambient temperature. After the establishment of a viable and thin biofilm (~ 100 µm) on the surface of Z-carriers, the influence of the changes of the organic loading rate (OLR) on the pseudo-first order degradation constants (kbiol) of MPs was evaluated in steady-state condition (Chapter (II)). The results revealed that Diclofenac, Naproxen, and 4n-Nonylphenol were biodegraded mainly by the biodegradation mechanism of co-metabolism, whereas the biodegradation of 17ß-Estradiol could be under the control of the mechanism of competitive inhibition. Individual contributions of the biofilm and suspended biomass on the abiotic and biotic removal of MPs were then investigated. In the case of abiotic removal, neither photodegradation nor volatilization could remove MPs, thereby abiotic removal of MPs was attributed to the sorption onto the biosolids. In this context, Naproxen, Diclofenac, 17ß-Estradiol and 4n-Nonylphenol, arranged in the ascending order of hydrophobicity, abiotically removed by 2.8%, 4%, 9.5% and 15%, respectively. In this regard, sorption of MPs onto the suspended biomass was seen around two times more than the biofilm. When comparing abiotic and biotic aspects, biotic removal outperformed its counterpart for all pollutants as Diclofenac, Naproxen, 17ß-Estradiol and 4n-Nonylphenol were biodegraded by 72.8, 80.6, 84.7 and 84.4%, respectively. kbiol values of all MPs were also seen higher in the biofilm as compared to the suspended biomass, especially for the recalcitrant Diclofenac.
In another part of the project (Chapter (III)), we aimed at determining whether bacterial bioaugmentation of tertiary MBBRs could successfully enhance MPs removal. The bacterial strain used for bioaugmentation was “Pseudomonas fluorescens”, that has a proven capability in both aspects of the biofilm formation, and in metabolizing the industrial pollutants. Two out of three MBBRs were inoculated by P. fluorescens with a novel protocol, and operated under the identical condition with the non-bioaugmented (control) MBBR (cMBBR). From the results of the DNA extraction and qPCR, the abundance of P. fluorescens in the biofilm and liquid phase declined with time. Despite this, bioaugmented MBBRs (bMBBRS) showed higher kbiol (pseudo-first order degradation constant) values than the cMBBR for all target MPs, along with a wonderful biotic removal i.e. 84.5, 90.4 and 95.5% for Diclofenac, Naproxen and 4n-Nonylphenol, respectively. On the contrary, MPs sorption onto the biosolids declined after the bioaugmentation as the above compounds were abiotically removed by 0.4, 1.1 and 3.9%, respectively. As compared with bMBBRs, a higher abiotic removal (2.8-15%) along with only an about 10% lower biotic removal was seen in the cMBBR. Achieving the high level of biotic removals in the cMBBR is might be due to the well-performed adaptation process. If biomass is not well adapted to target MPs, the distance between the efficiency of bMBBRs and cMBBR will be probably higher than what was obtained. Despite the fact that bMBBRs showed a high potential for the elimination of target MPs (in particular Diclofenac), this technology still needs further detailed research to overcome existing challenges, such as increasing the survival and maintenance of the inoculated strains.
As a whole, a high level of MPs removal is achievable in tertiary MBBRs, leading to convert them to a powerful technology with supporting both bio-routes of co-metabolism and competitive inhibition., and also abiotic abatement. Troubleshooting and optimization of the bMBBRs seem a proficient approach for future studies to take a step towards the complete elimination of MPs.
- PEM-based NF membrane
PEMs are prepared by alternately adsorbing the oppositely-charged polyelectrolytes onto the supports using a layer by layer (LbL) technique and can serve as re-generable surface coatings with controllable physicochemical properties (e.g. surface charge, hydrophilicity, and thickness). By such a technique, PEMs of two weak polyelectrolytes poly(allylamine hydrochloride) (PAH) and poly(acrylic acid) (PAA) were coated on the surface of ultrafiltration (UF) supports to obtain PEM-based NF membranes. Two types of UF supports: hollow fiber silica (HFS) (Chapter (IV)) and flat-sheet polyacrylonitrile (PAN) membranes (Chapter (V)) were used for the surface modification. In the current thesis, a special emphasis was devoted to the use of a weak PEM-based NF membrane as an easy to clean membrane with a low salts rejection and a high MPs removal from secondary-treated wastewater.
Before starting the filtration experiments, desired numbers of (PAH/PAA) bilayers were coated onto the model surface (plasma-treated silicon wafers) to optimize the coating conditions (pH and ionic strength) and to investigate the buildup behavior and hydration of multilayers, something that cannot be precisely monitored on the membrane itself. The UF supports were then coated with the optimized PEMs by dip-coating method, and tested by several parameters such as permeability, salts and MPs rejection. In the case of modified PAN membranes, the PEMs were also post-treated by the thermal and/or salt annealing, and were carefully characterized before and after annealing by the above parameters. After filtration of MPs-bearing wastewater, sacrificial cleaning of the fouled membrane was also examined.
As demonstrated in Chapter (IV), (PAH/PAA)6 multilayers prepared at lower ionic strength (5mM NaNO3) showed a lower hydration and consequently a better retention of salts and MPs than PEMs prepared at higher ionic strength (50 mM NaNO3). Before saturation of the membrane, the apparent rejection of the hydrophobic 4n-Nonylphenol was the highest, followed by Diclofenac and then Ibuprofen and Naproxen. This gives a strong indication that hydrophobic interactions dominate the apparent rejection, with more hydrophobic MPs adsorbing more to the membrane surface. Once saturated, a reduction in the level of MPs rejection was seen as the role of hydrophobic interactions faded-off. In this regard, correlation between steady-state rejection of MPs and their relevant molecular weights showed compounds of larger molecular weights are relatively better rejected, indicating rejection on the basis of size exclusion. Also, a strong relationship seen between the steady-state rejection of charged MPs and their relevant minimum projection area (MPA) was an indication for the importance of spatial dimensions in their ultimate retention. In contrast to existing high-efficient commercial NF membranes that can retain both salts and MPs to a high extent, we could prepare a membrane with a very low salt retention (NaCl ~17%) combined with a very promising removal of MPs, with Diclofenac, Naproxen, Ibuprofen and 4n-Nonylphenol being removed up to 77%, 56%, 44% and 70% respectively. Low rejection of salts leads to the production of a low saline concentrate, something that will facilitate its biological treatment. Additionally, such membranes do not noticeably disturb the salinity balance of the effluent, making the filtered effluent much more appropriate for irrigation water.
The influence of PEMs’ post-treatment (thermal and salt annealing) on the properties and performance of the membranes was evaluated in Chapter (V). Although PEMs became more compact and less hydrated by thermal annealing, no improvement was observed for the ions rejection. Upon salt annealing at various salt concentrations, the highest ion rejection was observed for (PAH/PAA)15 membranes annealed in 100 mM NaNO3, interestingly without any decrease in the water permeability. MPs retention simultaneously with contact angle variations of such membranes was in-depth studied over a filtration time of 54 h. As the filtration continued until the membranes saturation, an increase in membranes hydrophilicity was observed, and like our previous findings, the role of molecular and spatial dimensions emerged in MPs rejection. The steady-state rejection of MPs in salt-annealed membranes was higher than the non-annealed counterparts (52-82% against 43-69%), accompanied with still low NaCl retention (~25% against ~17%). Additionally, we proved that such membranes could be easily cleaned using a sacrificial layer approach. The fouled membranes were cleaned by a cleaning solution to release both the foulants and the sacrificial PEMs coating, without employing any shear forces. Such a finding can be an environment-friendly approach as the energy used for a conventional back-washing is avoided.