Bench Scale Tests Reveal Impact of Temperature on Nitrogen Removal in Post Aerobic Digestion-Acclimated Biomass

Background
Several processes are involved in the nitrogen cycle – a key process for wastewater treatment is the nitrification process. Nitrification is a two-step process where ammonia oxidizing bacteria and archaea, AOB and AOA respectively, oxidize ammonium (NH4+) to nitrite (NO2-) while the final step is carried out by nitrite oxidizing bacteria (NOB) with the oxidation of NO2- to nitrate (NO3-). AOB and NOB are usually the main key players for NH4+ oxidation to NO2-. Different parameters have been proven to influence nitrification and these include pH, temperature, dissolved oxygen, nitrification inhibitors and inorganic carbon concentration (Lin et al., 2009). Post-aerobic digestion (PAD) following anaerobic sludge digestion is an interesting and not well-explored process to achieve removal of nitrogen via nitrification and denitrification (Novak et al 2011). The Trinity River Authority (TRA) of Texas Central Regional Wastewater System (CRWS) Treatment Plant is currently developing the preliminary design for conversion of a Digested Sludge Storage Tank (DSST) to a PAD reactor. The proposed PAD reactor will follow mesophilic anaerobic digestion of a thermally hydrolyzed pretreated sludge. Currently, there are three PAD reactors in operation worldwide, reducing total inorganic nitrogen (TIN) recycle streams at all operating facilities. The three existing PAD reactors receive mesophilic anaerobically digested effluent, whereas the CRWS PAD reactor would potentially be the first to receive a THP mesophilic anaerobically digested influent. Limited information is available on the impacts of dissolved oxygen, temperature, and inorganic carbon limitations on nitrification and denitrification in PAD. While nitrification and denitrification are well-established processes, further studies at the bench scale level could help shed light on critical parameters such as for example temperature. This will help guide future proper design of PAD facilities.
Objectives
1) Determine critical impacts of temperature on nitrification
2) Evaluate bacterial rates under challenged conditions
3) Establish effects and quantify N removal in PAD systems under variable conditions
4) Explore possible alternative losses of inorganic N (modeling under development)
5) Provide guidelines and suggestions for operational key parameters
Methodology
The batch tests were carried out using PAD biomass from the Boulder Wastewater Treatment Facility (WTF) and the Denver MWRD NTP. Biomass from both plants was preserved at 4°C and warmed up/acclimated to testing temperature in the water bath before beginning experimentation. The initial tests were carried out with undiluted PAD samples. Each biomass sample was discarded after the daily use to avoid large community composition changes over time. The first set of experiments focused on temperature impacts on nitrification. These tests were conducted in the water bath at the following temperatures: 24.5, 30, 40 and 50°C. The pH value was recorded on a 2-4minute interval. The lime spikes were initially based on the alkalinity required for 200 mg N/L of ammonia oxidation within a three-hour time frame.
Findings
Nitrogen Removal Comparison The Denver and Boulder PAD sample batch tests were conducted in parallel reactors to compare nitrogen removals between the two reactors. The PAD reactors operate in a similar fashion with differences that may impact overall bacterial populations. The Boulder solids retention time (SRT), average of 5 days, is operated significantly lower than the average Denver PAD SRT of 20-30 days. A comparison of the ammonia biomass removal rates is summarized in Figure 2. Generally, the specific ammonia nitrogen removal rates are similar between each of the reactors except for the results at 50°C. A similar summary of nitrate and nitrite nitrogen was also observed. The highest ammonia specific rate was achieved at 40C for Boulder and between 40-50C for Denver. Review of the nitrate and nitrite nitrogen specific rates shows a similar comparison of the two PAD biomasses. The Denver 50°C ammonia nitrogen high removal rate was not associated with nitrate or nitrite accumulation suggesting ammonia nitrogen removal at the elevated temperature can be related to another mechanism such as ammonia stripping. Ammonia oxidation rates increase with increasing temperature through 50°C, with 40 and 50°C approximately equal (Figure 2). The Denver NTP PAD biomass did not indicate high levels of nitrite oxidation, suggesting absence of a significant NOB community or extremely low relative activity. Indicated nitrite oxidation rates are shown as not applicable with most rate tests indicating 0 or <1 mg N/L change. Nitrite accumulation rates are 20-40% that of observed ammonium oxidation rates. A TIN mass balance was tracked throughout each batch tests along with percent difference, the summary results are in Figure 2. Similarly to Denver, Boulder's biomass ammonia oxidation rates increase with increasing temperature through 40°C, 50°C indicating a significant reduction to nitrification rates (Figure 3). The Boulder WTF PAD biomass indicated low levels of nitrite oxidation, suggesting suppressed NOB community or very low relative activity. Nitrite accumulation rates are 20-40% that of observed ammonium oxidation rates. The TIN mass balance was tracked throughout each batch tests along with percent difference, the summary results are presented in Figure 3.
Microbial Ecology Analysis Relative bacterial abundance was also examined. One of the observed differences that indicate a difference in the AOB and NOB populations between Denver and Boulder is the nitrate specific rate. While both nitrate specific rates were low, the Denver biomass indicated a rate of zero or near zero suggesting a significant out selection or suppression of NOB activity (Figure 4). The comparison of AOB relative abundances between the Denver and Boulder biomasses indicate over 2x higher abundance for the Boulder biomass. This higher abundance could be driven by a lower SRT, requiring increased AOB populations for an equivalent specific rate. Whereas the Denver SRT is approximately 3X that of the Boulder PAD reactor which can decrease overall abundance. The anaerobic communities are also of note in the PAD reactors. The Denver PAD biomass is dominated by fermenting bacteria, with almost 20% of the overall biomass being fermenting bacteria (Figure 4). This is three times the fermenting bacteria population of the Boulder biomass and is likely a result of the long SRT in the Denver system. Both PAD systems also had a significant sulfate reducing population present.
Conclusions
Bench-scale testing indicates that biological ammonium oxidation accounts for the majority of ammonium removal in PAD systems. However, it appears that ammonia off gassing and nitrous oxide emissions (N2O) are also critical pathways. These gaseous emissions are currently being evaluated with process modeling. For the denitrification pathway and based on the 16S DNA analysis, the PAD reactors appear to be AOB dominated with few NOB present. This suggests the presence of shortcut nitrogen removal processes. Further research will shed light on gaps including the fate of ammonia and potential N2O production with a focus to capture dominant pathways and critical parameters affecting the emissions.