McIntosh, Lilian

Lilian McIntosh is a speaker at WEFTEC 2026.

Titles from this speaker

Microbial Adaptation to Low DO Biological Nutrient Removal

Abstract

Introduction Biological nutrient removal (BNR) in wastewater treatment (WWT) requires aeration for biological processes, such as nitrification and phosphorus (P) removal. Low dissolved oxygen (DO) BNR is a means of maintaining effective WWT and operational capacity while reducing the amount of aeration (and other resources) required as the operating DO is decreased. It is also a critical enabler for more advanced shortcut nitrogen (N) removal technologies and for improving the reliability of biological P removal. However, this field of study lacks a fundamental understanding of microbial metabolic kinetics under low DO conditions. To address this knowledge gap, pilot work was performed to investigate microbial kinetic adaptation to low DO while also monitoring key operational parameters including N removal efficiency, simultaneous nitrification-denitrification (SND), enhanced biological P removal (EBPR), settling, and filamentous bacteria growth. Methods At Hampton Roads Sanitation District's Virginia Initiative Plant (Norfolk, VA), primary effluent was fed to a continuous-flow BNR pilot consisting of an anaerobic selector followed by 5 aerobic tanks-in-series at a rate of 5.5 m3/d. Throughout the study, the DO concentration in the aerobic zone was incrementally decreased from 2 mg/L to 0.2 mg/L. Both aerobic solids retention time (SRT) and temperature were maintained consistently at 8.5 days and 20oC, respectively. Twenty-four-hour composite samples from the primary and secondary effluent were analyzed daily for COD, NHx, NO2, NO3, TKN, OP, TP, VFA, TSS, VSS, SVI, and total alkalinity. Nutrient profiles across the BNR reactors were completed twice a week and filamentous bacteria count was obtained via microscopic examination once a week. Sensors within the BNR configuration included: DO, pH, TSS, and ammonia ion selective electrode. Declining DO (DDO) and SND-P uptake (SND/P) batch tests were used to investigate changes in microbial kinetics. The DDO test provides a measure of the respiration rate of nitrifiers as DO in the batch declines to zero. The DDO test yields an oxygen half-saturation coefficient (Ko) and maximum growth rate (µmax) for ammonia oxidizing bacteria (AOB), nitrite oxidizing bacteria (NOB), and endogenous respiration. The SND/P test measures substrate (N and P) uptake rates at various DO set points to obtain nitrifier Ko and µmax, denitrifier oxygen half-saturation inhibition coefficient (Kio), polyphosphate accumulating organism (PAO) Ko, and maximum P uptake rate. Results The pilot achieved stable nitrification at each DO set point which resulted in an average ammonia removal of 99 +/- 2%, additionally 92 +/- 9% of P was removed on average (Figure 1). Nitrifier adaptation to low DO was observed in both in-situ measurements and batch tests. Nearly 100% ammonia removal was achieved at each DO set point and the rate of ammonia uptake observed in-situ did not decrease meaningfully. The DDO test results indicate that the AOB Ko decreased as the pilot DO set point was reduced (Figure 2). The SNDP test also showed a decrease in the average estimated AOB Ko but the decrease was not significant (Figure 3). This demonstrates that nitrifiers were capable of achieving a higher affinity for oxygen throughout the adaptation period while the maximum ammonia removal rates remained relatively constant. OHOs also adapted to low DO conditions, with a bit of an unexpected result. A decrease in the Kio for denitrification was observed as the pilot transitioned to low DO (Figure 4). SND was expected to improve under low DO conditions, however, due to the apparent adaptation of heterotrophs, minimal additional SND was observed. The SND/P test revealed that, at all pilot operating DO set points, PAOs have an extremely low Ko. Repeatedly, at the lowest batch test DO set point of 0.1 mg/L, the P-uptake rate observed was at a maximum; thus the Ko of PAOs is expected to be less than 0.1 mg/L (Figure 5). Filamentous bacteria counts remained low and SVI30 was consistently ≤ 100 g/L throughout the transition to low DO (Figure 6). The results of the SVI30 and filament count measurements demonstrate that settleability can be maintained under low DO conditions, even in an A/O process with some degradable COD making its way to the aerobic zone. Conclusion The low DO BNR pilot successfully transitioned to low DO conditions while maintaining settleability, complete ammonia removal, and excellent EBPR. Throughout the adaptation period, four impactful observations have been made: 1. AOB Ko decreased, and in-situ ammonia removal rates remained consistent as operating DO was decreased from 2 mg/L to 0.3 mg/L, indicating nitrifier adaptation to low DO conditions. 2. OHOs similarly adapted to low DO conditions, showing a reduced Kio for denitrification as operating DO was decreased. Because of the decrease in Kio, no increase in SND was observed as the operating DO was decreased. 3. EBPR performed well across the range of operating DOs tested. Kinetic experiments show PAOs have an extremely low Ko for aerobic P uptake. 4. Settleability did not deteriorate as operating DO was decreased; SVI remained stable within an acceptable range and filament counts were relatively unchanged

This paper was presented at the WEFTEC 2024 conference in New Orleans, LA October 5-9.

SpeakerMcIntosh, Lilian

Presentation time

13:50:00

14:00:00

Session time

13:30:00

15:00:00

SessionMicrobial Ecology Selection Under Low DO Concentration

Session number409

Session locationRoom 253

TopicAdvanced Level, Energy Production, Conservation, and Management, Municipal Wastewater Treatment Design, Nutrients, Research and Innovation

Author(s)

McIntosh, Lilian, Morgan, Haley, McCullough, Kester, Gagnon, Alexandria, Klaus, Stephanie, Rauch-Williams, Tanja, Vanrolleghem, Peter, Bott, Charles

Author(s)L.A. McIntosh1, H.M. Morgan2, K.K. McCullough3, A. Gagnon3, S. Klaus3, T. Rauch-Williams4, P. Vanrolleghem5, C.B. Bott3

Author affiliation(s)1Virginia Tech / HRSD, VA, 2HRSD, VA, 3Hampton Roads Sanitation District (HRSD), VA, 4Metro Water Recovery, CO, 5Université Laval, QC

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Oct 2024

DOI10.2175/193864718825159570

Volume / Issue

Content sourceWEFTEC

Word count9

Description: Why does low dissolved oxygen favor biological phosphorus removal? Oxygen inhibition...

Why does low dissolved oxygen favor biological phosphorus removal? Oxygen inhibition of PAOs is an important part of the story

Abstract

[b]Introduction[/b] Bio-P is a microbial metabolic process performed by PAOs that requires cyclic anaerobic and aerobic conditions. Therefore, conventional bio-P in wastewater treatment (WWT) requires DO. For decades, guiding principles suggested that DO in aeration zones should be ≥ 2.0 mg DO/L to maintain optimal bio-P conditions (Narayanan et al., 2006). More recently, focus has shifted towards low DO BNR, as a means of maintaining effective WWT while reducing the amount of aeration (and other resources) required as the operating DO is decreased. Recent studies have suggested that low DO can result in effective and energy efficient WWT (Keene et al., 2017). However, this field of study lacks a fundamental understanding of the effect of low DO on bio-P. [b]Materials and Methods[/b] The continuous-flow WWT pilot was fed with raw influent wastewater from Virginia Initiative Plant (Norfolk, VA) and performed preliminary, primary, and secondary treatment. Secondary treatment consisted of an anaerobic selector followed by 5 equal-volume aerobic CSTRs-in-series, constituting an A/O BNR process. The BNR process received primary effluent at 5.5 m³/d, yielding a total hydraulic retention time of 7.2 hours. Throughout the 610 day study, DO was incrementally decreased from 2.0 to 0.2 mg/L, while aerobic SRT and temperature were maintained at 8.5 days and 20 °C. Batch tests were used to investigate microbial kinetics. The first test measured substrate (nitrogen and P) uptake rates at various DO set points to obtain K[sub]DO[/sub] for nitrification, denitrification, and P-uptake. The second test measured P-uptake at 10-14 DO set points from 0.03 to 6.0 mg/L to more accurately determine the PAO K[sub]DO[/sub]. [b]Results and Discussion[/b] At each DO set point, stable bio-P was achieved with an average OP removal of 94% (SD = 8%). Bio-P at low DO (0.2-0.4 mg/L) outperformed higher DO set points, showing lower median effluent OP and a narrower distribution, indicating greater stability ([b]Table 1[/b]). Batch testing revealed extremely low K[sub]DO[/sub] for PAOs at all pilot DO set points, suggesting that PAOs do not require acclimation to low DO. At the lowest batch test DO of 0.1 mg/L, the P-uptake rate was at a maximum, suggesting a PAO K[sub]DO[/sub] < 0.1 mg/L ([b]Figure 1[/b]). Unexpectedly, the P-uptake rate decreased at high batch test DO, a phenomenon observed consistently across all pilot operating set points. This prompted further investigation of P-uptake at high DO (1-4 mg/L) and ultra-low DO (0.03-0.1 mg/L). While the Monod equation is commonly used to model PAO oxygen kinetics, the observed high DO inhibition resulted in using the Andrews equation, an empirical Haldane-based model with an inhibition coefficient (K[sub]i[/sub]) reflecting substrate (in this case oxygen) inhibition ([b]Figure 2[/b]) (Andrews, 1968). Batch tests performed over this large DO range confirmed K[sub]DO[/sub] < 0.1 mg/L for nearly all conditions, including full-scale plant mixed liquor ([b]Table 2[/b]). In each test, the same phenomena were observed: decreasing P-uptake rate as the DO of the batch increased. The impact of high DO was not a result of adaptation, but rather, an instantaneous effect. Due to the observed inhibition, the Andrews model consistently provided a significantly better fit (P < 0.05) than the Monod model. In this case, the K[sub]i[/sub] term in the Andrews model reflects the highest oxygen concentration at which the max P-uptake rate is equal to half of the max P-uptake rate without inhibition (Andrews, 1968). The resulting K[sub]i[/sub] was much greater at high DO than at low DO. This suggests that PAOs, which are normally operated at low DO, may experience greater susceptibility to inhibition when exposed to high DO. [b]Conclusion[/b] Results from this pilot study indicate that low DO improves bio-P, that PAOs have an extremely low K[sub]DO[/sub] (< 0.1 mg DO/L), and that PAOs P-uptake may be inhibited by high DO concentrations. PAOs did not need to kinetically adapt to low DO conditions as their K[sub]DO[/sub] was consistently low (<0.1 mg DO/L), regardless of pilot operating DO, as observed in batch tests. Batch testing also revealed a completely unexpected (and immediate) decrease in P-uptake rate at high DO, suggesting some form of oxygen inhibition. This finding served as the basis for employing the Andrews equation to model the collected data. The Andrews model provided a significantly (P < 0.05) better fit than that of the Monod. The idea that PAOs may be inhibited by high DO directly challenges the long-standing assumption that higher aeration ensures better bio-P performance. Evidence from both pilot and full-scale testing suggests that oxygen inhibition occurs at high DO, underscoring the advantages of operating at low DO. This study demonstrates that low DO BNR can simultaneously enhance bio-P and reduce effluent OP variability. As such, low DO bio-P represents a transformative and sustainable strategy for BNR, providing more reliable performance while conserving resources.

This paper was presented at WEFTEC 2026 in New Orleans, Louisiana.

Presentation time

16:30:00

16:45:00

Session time

15:30:00

17:00:00

SessionAdvancing Phosphorus Removal: What We Still Have to Learn

Session locationErnest N. Morial Convention Center

TopicFacility Operations and Maintenance, Municipal Wastewater Treatment Design, Nutrients, Research and Innovation

Author(s)

McIntosh, Lilian, McCullough, Kester, Gagnon, Alexandria, Klaus, Stephanie, Bott, Charles

Author(s)L. McIntosh1, K. McCullough1, A. Gagnon1, S. Klaus1, C. Bott1

Author affiliation(s)Hampton Roads Sanitation District (HRSD), 1Hampton Roads Sanitation District (HRSD), 1Hampton Roads Sanitation District (HRSD), 1Hampton Roads Sanitation District (HRSD), 1Hampton Roads Sanitation District (HRSD), 1

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Sep 2026

DOI10.2175/193864718825160353

Volume / Issue

Content sourceWEFTEC

Word count21

Base text size -

McIntosh, Lilian

Site admin