Investigating the effect of magnetisable glass foam particles (MGFP) in an UASB reactor treating synthetic winery wastewater by monitoring biofilm development and activity of colonised MGFP

[摘要] ENGLISH ABSTRACT: The growing of grapes for wine production is a generational, agricultural practice that has been associated with generating large revenue in South Africa. However, due to one of the outcomes of wine production, i.e. the generation of heavily contaminated wastewater, wine industries often face the obligation to treat wastewater prior to discharge to the municipality or irrigation. In addition, with the pressing matter of water scarcity at hand in South Africa, stricter regulations have been imposed on the treatment of winery wastewater (WWW), considering that wine production generates bulk volumes of wastewater. Since this winery wastewater is partially acidic and is characterised by high amounts of organic matter, inorganic ions, total suspended solids and polyphenols, these substances contribute to the pollution of water. Therefore, it is crucial for WWW to be depolluted to the standard specifications lest it negatively disrupts the ecosystem upon its discharge. Among the winery wastewater treatment methods developed and used in the wine industries, biological treatment methods (e.g. anaerobic digestion) are regarded as the most effective in treating winery wastewater (80 - 98% efficiency). The Upflow Anaerobic Sludge Blanket (UASB) bioreactor, which was used in this study, is one of the successful biological methods widely used at a lab-scale and commercial scale. The UASB reactor is primarily centred on the breakdown of organic matter to produce methane, a source of energy. However, the operation of UASB reactors often has the problem of sludge washout, which, consequently, results in reactor performance deterioration. Hence, in an attempt to prevent this problem, biofilm carrier particles known as magnetisable glass foam particles (MGFP) were used in this research. The study was focused on comprehensively investigating the effects of MGFP in an UASB reactor treating synthetic winery wastewater (SWWW) and monitoring biofilm development and activity of the colonised MGFP. The SWWW, which mimics the industrial winery wastewater, was used to make the substrate. The study was divided into two phases, whereby the first phase was aimed at treating SWWW at a gradually increasing organic loading rate (OLR) ranging from 0.5 to 5.0 kgCOD.m-3.d-1. Phase 2 was also focused on treating SWWW at a constant OLR (5.0 kgCOD.m-3.d-1), which resembles the common industrial OLR used in wine industries. Phase 1 was characterised by a high treatment efficiency with an increase in OLR. During Phase 1, from day 0 – 107, the COD reduction ranged from 71.4 to 97.7% in both Rcontrol and Rmgfp. The alkalinity, which indicates the strength of the buffer system, ranged from 450 to 3 075 mgCaCO3.L-1 in both Rcontrol and Rmgfp. The pH and concentration of volatile fatty acids (VFA), which also determine reactor stability and performance, were also within the standard specifications. The average VFA concentration in Rmgfp and Rcontrol ranged from 25 – 425 mg.L-1, which was within the optimal standard of < 500 mg.L-1 in both reactors, while the average pH ranged from 7.48 – 8.40 in Rcontrol and Rmgfp. There was a stable production of a high methane biogas by both reactors (> 55% methane), which ranged from 63 – 74% in Rcontrol and Rmgfp. The concentration of total suspended solids (TSS) measured in the effluent gradually increased with increase in OLR. The TSS increased from 100 - 380 mgTSS.L-1 in Rcontrol and from 80 – 400 mgTSS.L-1 in Rmgfp. Nevertheless, towards the end of the Phase 1, there was a reactor performance disturbance due to sludge washout caused by a high biogas production and this reduced the treatment efficiency of the reactors. Optimal reactor performance was restored in Phase 2 (day 108 - 180) due to improved settling of the sludge bed, moreover, due to stable operation of the reactors. During Phase 2, the water quality parameters were within the optimal standards for operating the UASB. The average COD reduction, alkalinity and pH ranged from 71.4 - 97.7%, 1 500 – 2 750 mgCaCO3.L-1 and 7.53 - 8.25 in Rcontrol and Rmgfp.The average TSS concentration in both reactors reduced from 380 to 360 mgTSS.L-1 in Rcontrol and 720 to 160 mgTSS.L-1 in Rmgfp. When Rmgfp, Rcontrol, control granules and colonised MGFP were analysed with a scanning electron microscope (SEM), a dense biofilm coverage was observed from the second month until the sixth month of the UASB operation. Both cocci and rod-shaped bacteria were observed in all of these samples, except for the control MGFP. In addition, the presence of bacteria and methanogens was corroborated under the fluorescence microscope where normal bacteria were distinguished from methanogens. The normal bacteria illuminated green while methanogens were blue as they auto-fluoresce. After performing granule activity tests, the results noted indicated that, overall, the colonised MGFP had the highest biological activity and acidogenic activity. This was presumably as a result of the presence of iron in the particles that aid in biogas production. However, Rcontrol granules had the highest biological activity when an acetic acid media was used, thus presumably suggesting that the sample had the highest population of active acetoclastic methanogens. Although, the magnetisable particles had negligible effects on the treatment efficiencies of the reactors, overall, the incorporation of the MGFP in an Rmgfp reactor had a positive impact, as an active anaerobic biofilm attached to the particles and produced a higher methane biogas. More so, due to the magnetic properties of the MGFP it was also feasible to extract them with a magnetic rod so that they could potentially be used as a source of multiplying active biomass to either seed another treatment process or be stored for cases of emergency (i.e. reactor failure or loss of biomass).

[发布机构] Stellenbosch University