Hydrolysis and Carbon Utilization for Low Energy Biological Nutrient Removal

Introduction and Background
Carbon transformations and bioavailability impact overall nutrient removal as well as sludge settleability. Bioavailability of organic carbon is a known limiting factor for simultaneous N and P removal. A significant fraction of organics in municipal wastewater is in the particulate and colloidal form (slowly biodegradable COD, sbCOD) that must be hydrolyzed to readily biodegradable COD (rbCOD) before it can be utilized by microorganisms. Hydrolysis is the rate limiting step in denitrification, as the process is slower than heterotrophic growth (Inset et. al, 2003). For simultaneous nitrification and denitrification (SND) systems, hydrolysis of particulate COD during aerobic zones can increase the carbon availability for denitrification. Prior to aeration, some hydrolysis will also occur within anaerobic selectors, where sbCOD can be converted into intracellular carbon storage, such as poly-b-hydroxybutyrate (PHB). In subsequent aerated zones, internally stored carbon can be utilized for denitrification. Aerobic hydrolysis rates can be measured using batch or semi-continuous respirometry methods and curve fitting to surface-saturation-type rate equations (Inset et. al, 2003). These methods are able to estimate the maximum hydrolysis rate for sbCOD when oxygen is not limiting. However, these rates do not reflect observed hydrolysis rates at low DO or anaerobic conditions, which are favorable for anaerobic uptake of carbon and for SND. As part of a Water Research Foundation study (WRF 5083), laboratory methods are being developed to estimate hydrolysis rates under aerobic and anaerobic conditions. Hydrolysis coefficients have been estimated using respirometry (oxygen uptake rate (OUR)), and curve fitting (see Figure 1 by Insel et al., 2003). This method is being adapted for low DO and anaerobic conditions in the Sturm lab at KU, with the nitrate utilization rate (NUR) being used instead of OUR. Internally stored carbon, as PHB, is also being measured within sludge acclimated to low DO BNR conditions. Experimental rates for hydrolysis under anaerobic and low DO conditions will be presented along with modeling results that demonstrate how the hydrolysis rates in sidestream fermentation reactors affect the design for advanced BNR systems. Experimental rates are being determined from pilot reactors fed with primary clarifier effluent and acclimated to low DO setpoints (< 0.5 mg/L). The laboratory research contributes to the larger WRF 5083 project and will be used to evaluate model calibration of full-scale treatment plants.
Materials and Methods
Pilot Operation Two identical 200-L continuous flow pilot reactors were operated at a flowrate of 575 L/day at the Kansas River WWTP, KS, US. The pilots were fed with primary effluent wastewater that passed through two anaerobic selector zones and three aerobic zones prior to settling. Aeration and mixing were achieved with porous air stones and axial flow impellers to maintain a DO setpoint between 0.3 - 0.5 mg/L in the first train and 1.5 - 2.0 mg/L in the second train.
Experimental Hydrolysis Rate Determination Activity batch tests were used for evaluating hydrolysis rates: the conventional oxygen utilization rate (OUR) test for aerobic hydrolysis and the nitrate utilization rate (NUR) test for anaerobic hydrolysis (Ekema et. al, 1986). Oxygen uptake data was measured with a fiber-optic oxygen sensor (FireSting -PRO, PyroScience) for the aerobic OUR test. NUR tests were completed by adding 10 mg-N/L of nitrate while maintaining anoxic conditions, as observed by DO readings less than 0.1 mg/L. Then, the mixed liquor was spiked with a carbon source > 400 mg COD/L, so that the NUR values were at the maximum hydrolysis rates and were not affected by differences in rbCOD (Melcer, 2004). Nitrate concentration was observed over time with a nitrate sensor (1006 Nitrate ISE, YSI) and confirmed with ion chromatography analysis. Mixed Liquor Volatile Suspended Solids (MLVSS) measurements were completed according to standard methods (APHA, 1992.) Hydrolysis parameters are fitted for both aerobic and anaerobic hydrolysis according to the multiple experiment curve fitting procedure outlined by Insel et al., 2003 (see Figure 1).
Sensitivity Analysis of Hydrolysis Rates for Sidestream Fermentation (Modeling) In BioWin, aerobic hydrolysis of particulate substrate to readily degradable complex substrate is the product of a hydrolysis rate constant (2.1 d-1), heterotrophic biomass concentration, and a Monod expression for the ratio of particulate substrate to biomass COD. The hydrolysis rate is reduced under anoxic and anaerobic conditions by applying an efficiency factor of 0.28 and 0.04, respectively. However, based on modeling of conventional and sidestream EBPR processes, this model formulation has been re-evaluated and modified to a standard product inhibition model where the hydrolysis rate only decreases as rbCOD accumulates. That is, the hydrolysis rate is not reduced for unaerated conditions unless required for calibration purposes using an efficiency factor. The results from this work will be used to confirm this new model formulation and refine the hydrolysis kinetics including the product inhibition and efficiency factors.
Results
Aerobic hydrolysis parameters can be obtained from OUR respirometry batch tests. This method is being adapted for NUR. Initial results are presented in Table 1, with three carbon sources tested (nutrient broth, acetate, and potato starch), which represented different sbCOD concentrations. These initial tests were used to validate that anoxic conditions are maintained in the experimental set-up (Figure 2) and that increasing sbCOD concentrations yielded decreasing NURs, as expected. Continuing research will adjust the S/X ratios of complex wastewaters with a ratio of pCOD and rbCOD. Recycled activated sludge fed to a sidestream fermenter is also being tested as a model complex wastewater that undergoes hydrolysis. This wastewater, obtained from the Wakarusa, KS wastewater treatment plant, is typical of RAS fermenters being calibrated in the WRF 5083 study.
Conclusion and Significance
Aerobic hydrolysis is routinely modeled in process simulators, but recent efforts to calibrate sidestream fermenters for EBPR have required adjustments to the hydrolysis function. The results from this work will be used to confirm new model approaches for anaerobic hydrolysis.