Please use this identifier to cite or link to this item: http://hdl.handle.net/2440/70851
Type: Thesis
Title: Integrated membrane systems for the removal of cyanobacterial toxins and taste and odour compounds.
Author: Dixon, Mike
Issue Date: 2011
School/Discipline: School of Chemical Engineering
Abstract: Cyanobacteria are a major problem for the water industry as they can produce metabolites toxic to humans in addition to T&O compounds that make drinking water aesthetically displeasing. It is likely that this problem will be intensified by the effects of climate change through reservoir warming. Tropical cyanobacterial species are also becoming more prevalent in more temperate climates. The effective removal of cyanobacterial metabolites is therefore an increasingly important priority for the worldwide water industry. Integrated membrane systems (IMSs) provide a potential water treatment method for the removal of cyanobacteria and their metabolites. The four types of membranes that have been applied for drinking water treatment are microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse osmosis (RO). The main parameter controlling the use of membranes for a particular application is normally the pore size. In general, important factors that affect membrane filtration are; pore size of the membrane compared with the size of the particles or molecules, surface chemistry of the membrane (charge, hydrophobicity), flux, recoveries and the degree of fouling. Very few studies have investigated IMS for the removal of extracellular cyanobacterial metabolites and no studies are reported in the literature for these systems for the removal of cyanobacteria. In addition, no published literature is available on the removal of cylindrospermopsin or saxitoxin using membranes. In general, high molecular weight cut-off membranes such as MF and UF may be used to remove cyanobacterial cells and intracellular cyanobacterial metabolites, while low molecular weight cut-off membranes such as NF and RO of the appropriate size range have the potential to remove a range of extracellular cyanobacterial metabolites. For a water supplier to be confident of removing both intracellular and extracellular metabolites, an integrated system with membrane technology paired with another treatment (possibly another membrane) is required. This study assessed the application of IMS including UF and NF membranes, used in conjunction with coagulation and powdered activated carbon (PAC), for the removal cyanobacteria and their metabolites. The cyanobacteria Microcystis aeruginosa, Anabaena circinalis and Cylindrospermopsis raciborskii were studied, and the cyanobacterial metabolites cylindrospermopsin (CYN), microcystins, geosmin (GSM) and 2- methylisoborneol (MIB) were also investigated. A rapid bench scale membrane test (RBSMT) unit was used to test NF membranes for removal of the cyanobacterial metabolites, microcystin, CYN, MIB and GSM. Results indicate that NF fouling can be caused by the cyanobacterial metabolites themselves. Each of the cyanobacterial metabolites were effectively removed to greater than 90% on average by NF membranes with molecular weight cut-off (MWCO) < 300Da (tight NF). A lower MWCO NF membrane (loose NF) provided improved removal of the neutral cyanobacterial metabolites (CYN, MIB and GSM) over the 220hr test period which may have been due to pore fouling. Removal of MCLR decreased over the 220hr test period in some tests. The effect of fouling of the NF membranes and its impact on removal of cyanobacterial metabolites was studied in two treated waters of different NOM concentration, Palmer water (average dissolved organic carbon concentration, DOC = 3.5 mg/L) and Myponga water (average DOC = 5.8 mg/L). The flux decline and the metabolite removal was the same in both waters throughout the trial, indicating that fouling was similar in the two waters. The results of our study showed that tight NF membranes removed MIB and GSM to above 75% and microcystin analogues and CYN to above 90%. MCRR, MCYR and MCLR removal was 50-80% by the loose NF membrane. MCRR appeared to be the least well removed of the variants while MCLA removal was >95%. This result is surprising as MCLA has the lowest molecular weight of the four variants and MCRR has the highest. However, the net charge of MCLA is two and zero for MCRR. Interestingly, this result is supportive of the observation that MCRR is the most readily removed by activated carbon, and MCLA is the least, suggesting a size exclusion effect, possibly due to a higher hydrodynamic diameter of MCLA, for both processes (permeation through membrane pores and diffusion through activated carbon pores). This finding warrants further investigation but this was not within the scope of this study. While NOM concentration had no impact on the removal of most cyanobacterial metabolites, if a loose NF is selected to avoid NOM fouling then removal of cyanobacterial metabolites may be less effective. This study has shown that NF can be an efficient treatment process for the removal of extracellular cyanobacterial metabolites and can be effective as the final step in an IMS. In general the removals are as expected with respect to the molecular weight of the compound and the MWCO of the membrane; however, molecular structure can also influence removals. In particular, nanofiltration is an effective process for the removal of MCLA, a compound that is recalcitrant to activated carbon removal and is less readily oxidised by chlorine than the other microcystin variants. An IMS incorporating coagulation, PAC and UF was investigated for the removal of intracellular and extracellular cyanobacterial metabolites. Three species and their metabolites were studied: A. circinalis, C. raciborskii and M. aeruginosa. The impact of coagulation and PAC addition on the UF membrane flux was also studied. After testing the UF-IMS using Palmer WTP water, two natural blooms were used to further challenge the IMS. Similar cyanobacterial metabolite removals were observed for each feedwater. Cyanobacterial cells were completely removed using a UF membrane alone and when UF was used in conjunction with coagulation. Alum was the least effective coagulant for this purpose. Extracellular metabolites were removed by PAC addition; however, coagulation hindered this removal in some instances due to entrapment of the PAC particles within flocs. UF membrane fouling by NOM was reduced by coagulation and PAC addition. A specialized blended coagulant combined with PAC was the superior coagulant for reducing fouling of the membrane. The application of a UF-IMS is warranted as cyanobacteria can compromise water quality and treatment in many ways. IMS can provide a multi-barrier treatment approach which is favoured in many risk-based approaches including many global water safety plans.
Advisor: O'Neill, Brian Kevin
Chow, Christopher
Newcombe, Gayle
Ho, Lionel Shih Wei
Dissertation Note: Thesis (Ph.D.) - University of Adelaide, School of Chemical Engineering, 2011
Keywords: ultrafiltration; nanofiltration; cyanobacteria; microcystin; cylindrospermopsin; saxitoxin; 2-methylisoborneol; geosmin
Provenance: Copyright material removed from digital thesis. See print copy in University of Adelaide Library for full text.
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
01front.pdf292.6 kBAdobe PDFView/Open
02whole.pdf3.77 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.