Enzyme augmentation of a submerged anaerobic membrane bioreactor : effects on solids hydrolysis and membrane fouling
Teo, Chee Wee
Date of Issue2014
School of Civil and Environmental Engineering
Singapore Membrane Technology Centre
Anaerobic hydrolysis of particulates presents both a challenge and an opportunity for raw sewage treatment using membrane bioreactors. A significant portion of the total chemical oxygen demand of raw sewage is in the form of organic particulates. Prior to assimilation by the anaerobic consortia, these particulates have to be hydrolyzed, which tends to be the rate limiting step. Enzyme augmentation, in the form of hydrolases, may be a potential solution. By enhancing hydrolysis, the overall biotransformation rate can be increased, chemical energy in the organic solids can be efficiently tapped, and solids accumulation can be better managed. Furthermore, it can potentially lead to a higher biogas production rate and a lower hydraulic residence time requirement. Due to complete solids retention, anaerobic membrane bioreactors (AnMBRs) can be used to treat screened raw sewage directly, obviating the need for primary sedimentation and sludge digestion. However, slow hydrolysis rates under anaerobic condition may lead to significant solids accumulation in the bioreactors and consequently reduce their effective volumes. At high solids retention times typical for AnMBRs, solids accumulation tends to be further exacerbated. Therefore, it is proposed to supplement a mixture of crude hydrolases into AnMBRs, in an attempt to specifically enhance the hydrolysis of complex organics present in sewage. These complex substrates are typically constituted of proteins, carbohydrates and lipids in substantial proportions. Accordingly, crude hydrolases containing proteases, amylases and lipases were introduced in this study. An inherent benefit may be the potential retention of exogenous enzymes by the gel layers formed on the membranes, hence minimizing enzyme washout. It is also postulated that membrane fouling may be mitigated due to enhanced enzymatic hydrolysis of macromolecular foulants. This study is divided into several phases, which attempts to elucidate the effects of enzyme augmentation on the performance of an AnMBR treating raw sewage. Firstly, the optimal pH and temperature for the anaerobic hydrolysis of organic particulates in sewage were determined by statistical analysis. The ranges of pH, temperature and enzyme dosage studied were pH 5-9, 25-35 oC and 0-0.7% w/v respectively. Using full factorial design, the optimum conditions were determined as pH 7, 35 oC and 0.7% w/v of enzyme dosage. Higher temperatures and enzyme concentrations were found to be beneficial for enhancing anaerobic hydrolysis. Secondly, a series of batch experiments were conducted to elucidate the effects of varying enzyme dosages on sludge characteristics. Zeta potential, particle sizes, floc compactness, settleability and hydrophobicity are the physicochemical properties evaluated in this study. Additionally, the impacts on the concentrations of loosely and tightly bound extracellular polymeric substances (EPS), and soluble microbial products (SMP) were assessed. Pearson’s correlation test was adopted to statistically assess the linear correlations between these parameters. SMP protein, EPS protein, SMP carbohydrate, sludge volume index (SVI), fractal dimension (Df) and contact angle are the factors identified as having strong and linear correlations with enzyme dosages. Positive relationships of exogenous enzymes with EPS protein, SMP carbohydrate, and SMP protein may be attributed to the localization of exogenous enzymes within the EPS matrix and the enzymatic hydrolysis of bound EPS into SMP. During the SVI analysis, sludge stratification was observed at enzyme dosages of 0.33 g/g TSS or higher, indicating the reduction in settleability. However, Df and particle size distributions revealed insignificant changes in floc structures and sizes respectively, suggesting that deflocculation was not exacerbated. Increasing enzyme dosages resulted in higher sludge hydrophobicity, which could be contributed by the immobilized enzymes within the EPS matrix. Thirdly, the effects of enzyme augmentation on the biological performance of a laboratory scale AnMBR treating synthetic raw sewage were ascertained. By supplementing a hydrolase mixture at dosages ranging from 0.9 to 18 mL/g influent COD, average total and volatile suspended solids declined by approximately 19% and 22% respectively, compared to the control without enzyme augmentation. Contrary to initial expectation, insignificant improvement in COD removal from 94.5±0.5% to 95.3±0.6% was observed. Nevertheless, it is noteworthy that permeate quality was not compromised, suggesting the retention of exogenous enzymes within the AnMBR, which was corroborated by size exclusion chromatography. In general, average biogas production increased from 0.27 to 0.34 L/g influent COD, in spite of larger fluctuation attributed to perturbation of steady-state condition. Additionally, bound EPS concentrations declined and gave rise to higher SMP concentrations, suggesting the enzymatic hydrolysis of bound EPS into SMP, the latter possibly biomass associated. Low enzymatic activities were detected throughout the entire study, probably due to the instability of exogenous enzymes in the bioreactor environment. Fourthly, batch and continuous filtration experiments were performed to investigate the effect of enzyme augmentation on membrane fouling. The hydrolase mixture was either applied directly as free enzymes, or immobilized covalently onto virgin membranes. Fouling mitigation effect of the former was evident throughout the entire study, judging from the transmembrane pressure (TMP) profiles. However, the latter lost their effectiveness when operated for an extended period of 20 days, after which TMP reached a similar plateau value of ~40 kDa relative to the control. During the continuous filtration experiment, chemical and particle size analyses of the cake layers revealed the enzymatic degradation of the macromolecular foulants formed on the membranes and the agglomeration of bioparticles respectively. The former was further corroborated by confocal images of bio-cakes stained for proteins and carbohydrates, which illustrated low fluorescence intensities and surface coverages. The latter could increase the bio-cake porosity, therefore conferring lower cake resistance based on the Carmen-Kozeny equation. Lastly, indigenously produced hydrolases were recovered from anaerobic and aerobic waste sludges via ultrasonication and supplemented daily into the AnMBR. Enzyme recovery was optimized statistically using full factorial design, by varying sonication intensity (2.0-4.6 W/cm2) and duration (2-30 mins), resulting in the optimal condition of 3.3 W/cm2 and 30 mins. Under the proposed system, COD removal efficiencies improved slightly from 93.0±0.9% to 94.2±1.2% following the supplementation of anaerobic sludge extract, but declined to 91.5±1.8% when dosed with its aerobic counterpart. Excitation-emission matrix analysis of the aerobic sludge extract allows the identification of humic and fulvic acids-like substances as the culprit. Biogas production increased inappreciably from 2.0±0.2 to 2.1±0.6 L/d, but rose more substantially to 2.3±0.3 L/d when supplemented with anaerobic and aerobic sludge extracts respectively. In spite of significant biogas generation, a shortcoming of applying the aerobic sludge extract was the exacerbation of membrane fouling, possibly due to its high SMP content.
DRNTU::Engineering::Environmental engineering::Water treatment