Investigating the Use of Internally Stored Carbon in Post-Anoxic Denitrification

Three of Hampton Roads Sanitation District's (HRSD's) 5-stage conventional activated sludge Water Resource Recovery Facilities (WRRFs) add methanol as a carbon source for post-anoxic denitrification: the Virginia Initiative Plant (VIP), Nansemond Plant (NP), and Army Base (AB). From 2017-2020, the methanol dose at VIP averaged 0.50 lb COD/lb N removed across the treatment process, while much higher doses of 1.57 and 2.04 lb COD/lb N were required at NP and AB, respectively. Full-scale plant profile data at VIP indicates an estimated average of 33% of total NOx removal in the post-anoxic zone that cannot be attributed to methanol or endogenous decay (Figure 1). It was theorized that this remaining NOx removal is driven by anaerobically stored internal carbon, similar to or potentially the same as the storage of volatile fatty acids (VFA) as polyhydroxylalkanoate (PHA) by polyphosphate-accumulating organisms (PAOs) and glycogen-accumulating organisms (GAOs). The question explored here was the extent to which this internally stored carbon is carried through the pre-anoxic and aerobic zones unused and then contributes to denitrification in the post-anoxic zone. These results suggest that many factors affect this internally stored carbon-driven (internal C) denitrification, and it likely contributes to the significant methanol savings at VIP relative to NP and AB.
The capability for internal C denitrification was analyzed with mixed liquor collected from VIP, NP, and AB for lab-scale batch testing. These 'standard' batch tests included an anaerobic carbon storage phase spiked with acetate followed by aerobic and post-anoxic phases with and without external carbon addition (Figure 2). Internal C denitrification was observed in the batch reactors for mixed liquor from each WRRF, despite not being detected in all full-scale post-anoxic zones. The specific denitrification rate (SDNR) was measured after the mixed liquor was subjected to a 24-hour anoxic period to deplete any residual internal carbon, and these endogenous decay SDNRs were consistently lower than the standard batch test SDNRs when internal carbon is suspected to be present. This same temperature-controlled batch test was repeated over the course of several months for each WRRF, starting in late summer. The internal C SDNRs increased for the batch tests performed in winter, by as much as 1 mg N/g MLVSS/hr for VIP biomass (Figure 3). The ideal plug flow conditions in the batch test setting was potentially a reason for observing internal C denitrification with each[/i] mixed liquor; it occurs full-scale only at VIP, where the post-anoxic zone is much more plug flow-like. Specific factors affecting internal C denitrification were further evaluated through altering operation of the standard batch test, such as aerobic hydraulic residence time (HRT). Increasing the aerobic HRT by 2-4 hours lowered the SDNR by an average of 0.21-0.35 mg N/g MLVSS/hr. These results validate the importance of the aerobic phase operation to internal C denitrification, whether by controlling internal carbon availability or by some other mechanism. Biological phosphorus removal (bio-P) has been identified as a pre-requisite for internal C denitrification, likely because of the anaerobic storage phase (Vocks et al., 2005). Therefore, it was investigated whether or not mixed liquor from any[/i] bio-P plant was capable of internal C denitrification. Using mixed liquor from HRSD's bio-P plants that are non-nitrifying/denitrifying, no denitrification beyond that from endogenous decay was observed in the standard batch tests. Figure 4 shows an extended anoxic phase at the end of the test, and no change in the denitrification rate for the non-nitrifying WRRF biomass compared to a distinct change for VIP biomass when the internal carbon source presumably has been depleted. It was deduced that internal C denitrifiers need prior exposure to NOx, so bio-P biomass for this type of denitrification was not the only pre-requisite. Exposure to NOx is not a problem for WRRFs such as VIP, NP, or AB where nitrification/denitrification is obviously achieved, but it could factor into considerations as for aeration control that would affect NOx concentrations. These results also suggest that internal C denitrifiers are a distinct subset of heterotrophs. The effect of the amount of anaerobic acetate added in the batch test on the internal C SDNR was also evaluated, as anaerobic VFA augmentation has previously been shown to increase post-anoxic SDNRs (Coats et al., 2011). The increase in internal C SDNRs when the acetate dose was increased from 20 mg COD/L to 100 mg COD/L ranged from 0.06 to 0.28 mg N/g MLVSS/hr. Therefore, the quantity of carbon is one aspect of anaerobic storage that can control internal C denitrification.
Lastly, the internal C SDNRs were compared to the phosphorus (P) release and uptake rates for the batch tests. For each WRRF, there was a positive correlation between aerobic P uptake and internal C SDNR (Figure 5). There were no significant changes in the post-anoxic P concentration, so it is unlikely that denitrifying PAOs (dPAOs) are responsible for the internal C denitrification. Regardless, the trend with aerobic P uptake still indicates that internal C denitrification is tied to PAO activity. This is reflected in full-scale observations of the post-anoxic zone at VIP, where methanol addition and effluent P concentration tend to increase simultaneously (Figure 6). Higher effluent P concentrations from less effective bio-P performance can be expected in summer due to more competition from GAOs at higher temperatures (Lopez-Vazquez et al., 2009). If internal C denitrification is indeed linked to bio-P, poor bio-P could reduce contributions from internal C, and thus explain the higher methanol addition rate observed at these times. This dependence on bio-P could explain the increase observed in standard batch test internal C SDNRs during winter, because this is when bio-P performance is better at these WRRFs.
Understanding the conditions that promote internal C denitrification are required to develop effective strategies for full-scale implementation. For instance, the advantage of shorter aerobic HRT could be realized through lowering the aerobic DO setpoint. In that case, benefits could be two-fold to both external carbon and aeration energy savings. Another key aspect was the relationship between internal C denitrification and bio-P, so encouraging bio-P performance, such as through GAO out-selection, could improve not only P removal but nitrogen removal as well. The value of attaining internal C denitrification is exemplified by comparing methanol costs between WRRFs with and without this type of denitrification: while VIP treats almost three times as much flow as AB, their 2020 methanol cost was less than half of that of AB. A combination of operational changes that are advantageous to internal C denitrification based on the results of this study is likely to be a starting point for achieving methanol savings that would rival those at VIP.