Low-temperature anaerobic wastewater treatment by granulated biomass

Abstract

In this study, long-term operation of up-flow anaerobic sludge blanket (UASB) system treating real municipal wastewater at decreasing temperatures (25, 16, 12, 8.5, 5.5, and 2.5 °C) and variable organic loading rates from 1.0 gCOD·l⁻¹·d⁻¹ up to 15.2 gCOD·l⁻¹·d⁻¹ was investigated over 1025 days. Experiments were performed in two parallel in-house designed laboratory-scale UASB reactors, which were operated continuously with hydraulic retention time of 16.7 h down to 1.1 h. Stable COD removal efficiencies of 50 - 70 % were achieved at 25 °C down to 8.5 °C with loading up to approximately 15.2 gCOD·l-1·d-1. COD removal efficiencies were reduced at temperatures below 8.5 °C, but significant methane formation was observed even at 2.5 °C at reduced loading (up to 5 gCOD·l-1·d-1). More than 90% of COD removed was converted to methane, and the methane yield did not change significantly with respect to temperatures. The overall COD balance closed at above 90% of the inlet COD at all operating temperatures and organic loading rates. Temperature affected the reactor performances, microbial community structure, and the degradation pathway of organic matter with acetoclastic methanogen played significant roles. Acetate was the primary precursor of methanogenesis pathway at low-temperatures. Microbial communities proved the adaptation ability to very low-temperatures down to 2.5 °C regardless of the operating organic loading rates; psychrotolerant. Additionally, an anaerobic granulated biofilm system at 25 °C and different organic loading rates (1, 3, 8, 10, 15, and 20 gCOD·l-1·d-1) was modelled in AQUASIM 2.1 to predict and simulate biofilm model implementation and assumptions specific to the granules as a fixed biofilm in UASB reactor system in this study. Simulated organic loading rates scenario results showed COD removal efficiencies (62 - 69%) and methane fraction (83 - 88%) in biogas at steady-state conditions decreased with the increasing organic loading rates. All simulations predicted an increased pH profile, from pH 7 in the outer layer to approximately pH 8.3 in the core of granules, under increasing organic loading rates, but the biomass composition and active biofilm regions were not significantly affected by organic loading rate variable. UASB effluent post-treatment investigations, specifically on dissolved methane and nutrient removal, were performed as supplementary studies. A methanotrophic-cyanobacterial syntrophy was established in the existing oxygenic photogranules to remove dissolved methane. This syntrophy was maintained and propagated in a continuously operated reactor (hydraulic retention time of 12 h), proven by observed biomass yield and dissolved methane removal by approximately 2.4 gVSS·gCH₄⁻¹ and 85%, respectively, with COD balance closed at around 91%. Community analysis suggested methanotrophs and phototrophs syntrophy, and the cross-feeding between photogranules of different community compositions, containing methanotrophic bacteria, phototrophs, and non-methanotrophic methylotrophs. Nutrient removal from filtered UASB secondary effluent was investigated using several microalgal strains based on a literature review: Chlorella vulgaris, Chlorella sorokiniana, Tetradesmus obliquus, Haematococcus pluvialis, and Microchloropsis salina. Microalgae strain C. sorokiniana presented the ability to grow in wastewater in all the tested culture conditions, suggesting high adaptability and viability of the strain in this specific type of filtered UASB secondary effluent. The results also implied nutrient removal achieved 62% of total nitrogen removal and 97% of total phosphorous removal, when applying C. sorokiniana in the batch systems with hydraulic retention time of 9 days. The growth potential and the nutrient removal capacity of C. sorokiniana in a continuous laboratory-scale photobioreactor were also investigated. The system removed total nitrogen and total phosphorous by approximately 17% and 27%, respectively, with hydraulic retention time of 5.5 days. High ammonium removal yet high nitrate release indicated microalgae-nitrifiers imbalance in the photobioreactor system. Unfavorable growth factor for microalgae in the photobioreactor could be carbon-limited media in wastewater (2C:22N:1P). Adding an external source of CO2 to control alkalinity, pH, and provide carbon source for microalgal growth is most probably needed for further investigation. Even though nutrient removal efficiencies in continuous photobioreactor were significantly lower than the batch test, biomass yield in the photobioreactor was higher (0.16±0.02 gTNremoved∙gSS-1 and 0.03±0.005 gTPremoved∙gSS-1), compared to the batch test (0.04±0.01 gTNremoved∙gSS-1 and 0.02±0.002 gTPremoved∙gSS-1). High biomass production suggested that microalgal-based treatment for UASB effluent could offer a resource recovery potential. Overall, this study demonstrated the feasibility of UASB system treating municipal wastewater at low-temperatures and variable loadings in a long-term application. In combination with suitable post-treatments, UASB system showed a viable secondary pre-treatment option unit process for achieving lower carbon footprint wastewater treatment and resource recovery at low-temperatures. The robustness exhibited to low-temperature and variable loading conditions provides a solid basis for further research and potential applications. Further advances in an integrated UASB system and post-treatment unit processes investigation will be needed for pilot- and/or full-scale applications in the future.