Membrane bioreactors for sewage and stabilised leachate treatment

Abstract

The project aimed to extend the limited knowledge ofMBR (Membrane Bioreactor) operation for the treatment of stabilised (old) landfillleachate using an air-lift tubular membrane configuration and comparing the results with those obtained for sewage treatment under largely identical conditions. SRT (solids retention time) was used as the principal fixed variable in the two trials so as to allow comparison of process performance for the two different feedwaters. Supplementary tests were also conducted on: 1. bench-scale porous pots, used to identify the optimum HRT (hydraulic retention time) value for leachate treatment, and 2. intermittent aeration, used to minimise aeration demand. At bench scale removals of69 ± 1% and 99.9 ± 0.1% chemical oxygen demand (COD) and ammonia (NH/) respectively were achieved at an HR T of 5 d and temperatures ~ 21°C. For this trial the mixed liquor volatile suspended solids/mixed liquor suspended solids (MLVSSnvn.SS) ratio was 0.7 ± 0.03 and the volumetric loading rate (VLR) 0.4±0.02 kgCOD m- 3 dol and 0.15 ± 0.003 kgNlI/ m-3 dol; the COD and NH/ food to microorganism ratio (FIM) ratio values were respectively 0.1 ± 0.01 and 0.04 ± 0.004 dol; specific oxygen utilisation rate (SOUR) was 16 ± 7 mg02.gVSS- 1 .h- 1 and nitrifiers:heterotrophs ratio was around 80:20. This compared with similar nitrifiers proportions (i.e. 70 to 80%) reported in the literature. During pilot trials COD removals at the stabilised period were consistently larger for sewage (69 - 83%) than leachates at the pseudo stability phase (28 - 56%), the latter being somewhat lower than values reported in the literature for full-scale plants. However, for both feedwaters maximum was obtained at VLRs between 0.99 and 1.2 kgCOD.m- 3 .d- 1 • NRt + removals were readily achieved with values ~ 88.6%. Biomass characteristics, namely particle size distribution (PSD), extracellular polymeric substances (EPS) concentrations and dynamic viscosity (11) were found to be similar for both sewage and leachate sludges, while soluble microbial products (SMP), MLVSS!.MLSS, F/M, SOUR, and heterotrophslnitrifiers proportions differed. most likely i due to impact of the feedwater character. Microbial dynamics and speciation were highly dependent on feed water quality showing that highly variable feeds such as leachate and sewage would develop inconsistent bacterial communities but analogous to each other. Constant feeds, on the other hand, would develop highly consistent bacterial community profiles. Nevertheless species richness or abundance of neither group (i.e. inconsistent and consistent communities) was significantly correlated to microbial foulants production such as SMP andlor EPS. Even further it was found that either microbial community would have no effect on COD or TOC (total organic carbon) levels of the treated feeds. Critical flux (Je) tests revealed fouling of leachate biomass to be more severe than that imposed by sewage biomass. and that fouling by the leachate biomass is predominantly attributable to the feedwater itself. No correlation between fouling and conventional biomass foulants (i.e. SMP and EPS proteins and carbohydrates) was evident, whilst a stronger correlation of fouling was shown with TOC of the SMP fraction. The ceramic membrane material tested was, as expected, more resistant to fouling than the polymeric membranes used. The highest le values on continuous air-sparging for sewage and leachate were 36 and 24 L.m"2.h"1 respectively. for polymeric membranes, and about 30 L.m"z.h"l during intermittent air-lift conditions for sewage, while absent for leachate. For ceramic membranes no le values were observed at continuous air-sparging. whilst no fouling was evident for either polymeric or ceramic materials challenged with leachate operating with intermittent aeration (0.5 -1 Hz) up to a flux (1) ofl3 and 44 L.m"l.h"l. Cylindrical geometry lumens were found to be more effective during air-lift operations than square lumens.