Guo, Hui

Guo, Hui

Hui Guo is a speaker at WEFTEC 2026.

Joined:

About Guo, Hui

Hui Guo is a speaker at WEFTEC 2026.

Guo, Hui

Hui Guo is a speaker at WEFTEC 2026.

Titles from this speaker

Description: Defining Best Operating Points of Biomass Densification for MBR and Clarifiers:...

Defining Best Operating Points of Biomass Densification for MBR and Clarifiers: Lessons Learned from Two-year Operation at Full-scale Plants

Abstract

Densified Activated Sludge (DAS) in Continuous Flow with MABR and MBR has been implemented at full-scale wastewater treatment plants for over 2 years. While new indicators such as SVI5, DI% and TTFOF have been revealed to be enhancing process control and helping operators better scrutinize and control process events (Shaw, Donnaz et al, 2024), steady-state operations also showed the existence of a climax for clarifier and ultrafiltration performance in terms of SVI30 and Densification Index. DI is defined as the sum of bio-aggregates > 200 µm in the Total Suspended Solids (gMLSS>200/gMLSS).

At Yorkville Wastewater Treatment Plant (WWTP), the long term operation of MABR DAS with clarifiers demonstrated the ability to operate with half of the hydraulic capacity of the plant (Regmi, Astrand et al, 2024; Roche et al, 2021; Donnaz et al, 2020).

From November 2023, SVI30 reached 72 mL/g at 95%ile and 58 mL/g on average. TSS in the outlet of the clarifier reached 13 mg/L at 95%ile and 5 mg/L on average (Figure 1).

However, when SVI30 reached a threshold below 40 mL/g, TSS in the secondary effluent increased from 5-7 to hit the 13 mgTSS/L mark. This was rapidly compensated by adjusting SVI30 back between 50 and 60 mL/g and by increasing the RAS rate from 45 to 50%.

Total SRT of Yorkville secondary biological system is 11-13 days, aerobic SRT 9-11 days, F/M 0.26 kgCOD/kgMLVSS/d (COD/BOD ratio 1.6).

At Detroit Lakes WWTP, the long term operation of MBR DAS with ultrafiltration provided benefits (Donnaz, Shaw et al, 2023) in terms of membrane resistance (-30%), colloidal fraction (-60% cCOD to -80% cTOC) and membrane cleaning (maintenance and recovery). Improvement of membrane filtration was correlated to rapid drops of SVI, time-to-filter and superfine colloids (sfCOD) in the membrane tank mixed liquor (Donnaz, Shaw et al, 2023). Time-to-filter is a direct test of filterability and membrane performance. Results of these tests show a very close correlation with settleability, meaning that improved settling and higher densification also improve filterability (Shaw, Donnaz et al, 2024). Total SRT of Detroit Lakes secondary biological system is 15-25 days, aerobic SRT 9-10 days, F/M 0.05 kgBOD/kgMLVSS/d.

This was confirmed in line with supporting research demonstrating that MBR mixed liquor quality can be enhanced by using hydrocyclones as gravimetric selectors, whereby MBR performance has been improved due to the removal in overflow of lighter flocs and colloidal fraction from the mixed liquor (Nogushi et al, 2018). External gravimetric selectors help regulate particles and granules size below 1 mm, which is believed to minimize pore blocking (Zhang & Jiang, 2018). Particle size distribution of densified biomass has shown to be 200 to 500 µm in majority, with 1,000 µm maximum (Roche et al, 2021). An external gravimetric selector such as memDENSE technology helps regulate particle and granule sizes above 100 µm and below 1 mm, which is believed to minimize pore blocking fouling observed with full granulation process by remaining below the 1 to 1.2 mm critical value range described by Zhang & Jiang (2018).

As DI% evolved from 30% to 80% in the biological system (Figure 2), the following was observed (Figure 3):
- optimum of TTF around SVI30 of 60 mL/g
- drop of permeability observed for SVI30 >70 mL/g and SVI30 < 40 mL/g
- optimum Permeability observed around DI 50%

DI-SVI correlation enables the prediction and the accurate assessment of DAS mixed liquor performance in secondary biological continuous flow systems (Guo, Donnaz et al 2024). Experimental data show that there is a linear correlation between DI and Particle Size Distribution (PSD) with DAS for all particle fractions.

Figure 4 shows how PSD evolves as DI increases over 50% in the system (aerobic and underflow):
- Fraction >550 µm increases when DI gets over 50% (blue)
- Fraction 200-550 µm is the one which increases the most when DI gets >50% (orange)
- Particles fraction < 100 µm (Σ<50;50-100 µm) decreases significantly as DI augment (yellow and red)
- Fraction 100-200 µm is the third that is decreasing the most after DI increases over 50% (green)
- Correlation between PSD and DI% shows that as DI increases from 40 to 80%, volume % of particles
> 200 µm increases while volume % of particles < 200 decreases

Evolution of DI above 50% in the biological system introduces the presence of big granules (>550 µm), while smaller granules are still abundant (relative vs absolute fractions). Therefore, there is a combined effect of the emergence of bigger granules (teal blue) close to the critical threshold of 1-1.2 mm and pore blocking by fine particles on the membrane permeability (Campo et al 2021). While Campo et al claim that one of the main factors for stable operation is to "verify that granules mean dimensions are far from critical values (1-1.2 mm)", these results imply a noticeable drop of permeability when d90 starts to flirt with 590 µm. It appears that the shear effect of the memDENSE selector's ability to self-regulate the size of the particles below 1 mm functions until DI reaches 60-80%.

Densified Activated Sludge in continuous flow has been coupled with membranes at full-scale with two different separators: secondary settling tank (MABR and clarifier) and ultrafiltration (MBR). In two projects, steady state densification has been implemented to find the best the best operating points.

This paper was presented at WEFTEC 2025, held September 27-October 1, 2025 in Chicago, Illinois.

Presentation time

16:00:00

16:10:00

Session time

15:30:00

17:00:00

SessionDensified Activated Sludge Full Scale Performance and Optimization

Session locationMcCormick Place, Chicago, Illinois, USA

TopicAdvanced Water Treatment and Reuse

Author(s)

Donnaz, Sylvain, Guo, Hui, Shaw, Chris, Astrand, Niclas, Gagnon, Jean, Fyfe, Sheila, Reeve, Matt

Author(s)S. Donnaz1, H. Guo1, C. Shaw1, N. Astrand1, J. Gagnon1, S. Fyfe1, M. Reeve1

Author affiliation(s)Veolia WTS USA Inc.1

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Sep 2025

DOI10.2175/193864718825159937

Volume / Issue

Content sourceWEFTEC

Word count20

Densification Index/SVI Model: An Efficient Way to Predict Benefits of Biomass Densification in Full-scale MBR and MABR Wastewater Treatment Plants

Abstract

Abstract Densification of biomass is an efficient technology to improve sludge settleability which benefits wastewater treatment plants (WWTPs) by increasing the capacity of the secondary clarifiers. In this study, a densification model was developed and validated to predict the level of densification performance to help capture expected benefits and new capacity limits for operations. The first version of the VWTS (Veolia Water Technologies & Solutions)-SUMO model was developed based on the growth kinetic for various microorganisms (e.g., nitrifiers and polyphosphate-accumulating organisms). Then the model was calibrated with the filed data collected from full-scale WWTPs (WWTPs with membrane-aerated biofilm reactors and WWTPs with membrane bioreactors). Although more filed data under steady-state operating conditions need to be collected to validate the model, the current simulation results demonstrated the settleability in the full-scale WWTP can be improved significantly by the densification processes. In addition, the proposed model can predict the settleability reliably. 1.INTRODUCTION Densification and granulation of biomass have been demonstrated to be efficient in improving sludge settleability which determines the design and performance of secondary clarifiers (i.e., the hydraulic capacity) in wastewater treatment plants (WWTPs). For instance, a full-scale granule-floc hybrid process in Dijon (France) proved that 20% of aerobic granular sludge (AGS) guarantees a sludge volume index (SVI) lower than 100 mL/g while 30% of AGS reduced the SVI to 50 mL/g [1]. The improved settleability (i.e., lower SVIs) further contributed to a better performance of clarifiers with two to three times higher surface loading rates (SLR) and surface overflow rates (SOR). However, the densification and granulation processes, configurations, and mechanisms still need to be understood more deeply. Thus, a densification model is needed to predict the level of densification performance (i.e., SVI) to help capture expected benefits and new capacity limits for operations. Densification Index (DI), which is the percentage of biomass larger than 200 µm in the mixed liquor, was proposed to quantify the level of biomass densification and partial granulation in continuous-flow biological treatment systems. In this study, a simple and reliable protocol was built to capture the DI in a full-scale WWTP with selective wasting (i.e., densification). These sites' DI and SVI data were used to implement and validate an SVI prediction model. With the predicted SVI from this model, the hydraulic capacity in WWTPs can be upgraded to reach new levels of process intensification. 2.METHODS 2.1 VWTS-SUMO Model The VWTS (Veolia Water Technologies & Solutions)-SUMO model [2][3] provides a simple method of calculating and predicting DI on the state variables from the process simulator. The model introduced separate growth kinetics for granular nitrifiers (i.e., ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB)), polyphosphate-accumulating organisms (PAOs), and ordinary heterotrophic organisms (OHOs) that contribute significantly to the formation of denser and granular biomass (> 200 µm). The half-saturation coefficient (Ks) of PAOs, AOBs, NOBs, and OHOs and their effects on granular growth were implemented based on the previous framework [4]. The contribution of influent inorganic suspended solids (ISS, > 200 µm) was also added to the model. 2.2 Field Data Collection and Calibration A simple and reliable protocol (i.e., sieve tests) has been implemented to measure the DI of a full-scale A2O (anaerobic-anoxic-aerobic)-MABR (membrane aerated biofilm reactor) WWTP and a full-scale A2O-MBR (membrane bioreactor) WWTP. The configuration of the treatment plant is shown in Figure 1. A SUMO model was developed and validated by the influent wastewater characteristics using VWTS blocks specially designed and developed for MABR, Secondary Settler, and External selector. RESULTS AND DISCUSSION The first version of the DI-SVI prediction model was proposed based on the VWTS-SUMO model (Table 1). The VWTS-SUMO model was re-calibrated by the collected data from the full-scale WWTPs. According to the site data, the model can reliably predict SVI before the densification process (Figure 1) for both MABR and MBR applications. However, modified equations should be used after the densification process since the SVI (i.e., settleability) in the full-scale WWTPs was improved significantly. For instance, the SVI can be reduced from 101 mL/g to 65 mL/g while the DI increases from 30% (typical DI in an aeration tank of CAS) to 50% in a densified MABR WWTP (Figure 2a). It should be noted that the densification process in this full-scale WWTP has not reached a steady state since the operating time for the densification process was around 7 months for the MABR plant and 5 months for the MBR plant. An updated model will be proposed by VWTS with more field data. In addition to the settleability improvement by the densification process, a modified VWTS-SUMO model will be developed to predict carbon, nitrogen, and phosphorus removal.

The study developed a densification model to predict the performance of biomass densification, a technology that improves sludge settleability and increases the capacity of secondary clarifiers in wastewater treatment plants. The model, based on the growth kinetics of various microorganisms, was calibrated with data from full-scale wastewater treatment plants.

SpeakerGuo, Hui

Presentation time

13:30:00

13:50:00

Session time

13:30:00

15:00:00

SessionMechanistic Modeling Developments for Newer Processes

Session number410

Session locationRoom 244

TopicAdvanced Level, Biosolids and Residuals, Industrial Issues and Treatment Technologies, Municipal Wastewater Treatment Design, Research and Innovation

Author(s)

Guo, Hui, Donnaz, Sylvain, Shaw, Chris, Astrand, Niclas, Houweling, Dwight

Author(s)H. Guo1, S. Donnaz2, C. Shaw3, N. Astrand4, D. Houweling5, H. Guo1, S. Donnaz2

Author affiliation(s)1Veolia WTS, Ontario, 2Veolia WTS, MT, 3Veolia Water Technologies and Solutions, Ontario, 4Veolia Water Technologies & Solutions, ON, 5Dynamita North America Inc., QC

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Oct 2024

DOI10.2175/193864718825159686

Volume / Issue

Content sourceWEFTEC

Word count21

Description: MABR Biofilm Technology Reduces N2O Emissions: Results from Eight Full-scale...

MABR Biofilm Technology Reduces N2O Emissions: Results from Eight Full-scale Mainstream & Sidestream Applications

Abstract

Background Nitrous oxide (N2O) is produced during biological nitrogen removal at wastewater treatment plants (WWTPs) (Kemmou, L. and Amanatidou, E., 2023). N2O emissions can contribute up to 75% of greenhouse gas (GHG) emissions from a WWTP (Daelman MR et al., 2013), therefore it is critical for facilities to mitigate N2O emissions to minimize their GHG footprints. Membrane aerated biofilm reactor (MABR) technologies have demonstrated greenhouse gas (GHG) reduction due to their high aeration efficiencies and have also shown great potential for N2O mitigations. However, studies on N2O emissions from MABR technologies were mainly carried out by modelling or using lab-scale systems (He and Daigger, 2023; Bunse et al., 2024). Measurement of N2O emissions from full-scale MABRs is rare (Uri-CarreÃ±o et al., 2024; Long, Z. et al., 2023). Although literatures indicated that MABR technologies can reduce N2O emissions, the extent of reduction may depend on many factors. In this paper, N2O emissions from eight full-scale MABR systems are reported with affecting factors examined. The main objectives include, (1) to establish the benchmark(s) of N2O emissions from MABR technologies; and (2) to improve the understanding of how MABR technologies can mitigate the N2O emissions. Methodologies Figure 1 shows the N2O monitoring techniques and the approach to determine the N2O emission factors. Table 1 provides a simplified description for the full-scale MABR plants in this study. Results and discussion 1. N2O emission benchmark for MABR unit process Table 2 lists all available N2O EFs from the study, with the data focusing on MABR unit process. Table 1 indicates that MABR exhaust is the main source of N2O in MABR unit process, with the mean N2O EF of 0.55%Â±0.65% in MABR exhaust. The total N2O EF from MABR unit process was slightly higher, with the mean N2O EF of 0.61%Â±0.90%. Comparing with the N2O benchmark from IPCC 2019 (EF of 1.6%), MABR unit process reduced N2O emission by approximately 2/3 to remove the same amount of NH4+-N. In addition, MABR exhaust can be easily collected and the final disposal of N2O in the exhaust would be much more cost effective due to the low volume of the exhaust. Following factors might have contributed to the low N2O EF from MABR tanks. â— Counter diffusion in MABR biofilm: N2O generated by AOB nitritation and de-nitritation in MABR biofilm will diffuse either across the membrane into the gas phase or across the outer biofilm into the bulk liquid. The outer biofilm might serve as a mass transfer barrier and even a N2O sink, resulting in low N2O transfer into the bulk liquid. â— Denitrification in the biofilm and/or the mixed liquor: the bulk liquid in MABR tanks is normally anoxic or anaerobic. In addition, carbon source in the feed is often available for the biofilm and suspended biomass if present. The heterotrophic denitrification by the biofilm and/or the suspended biomass would further minimize the N2O concentration in the bulk (Conthe, M. et al., 2019). â— Intermittent mixing in MABR tanks: the intermittent mixing would minimize the N2O emission into the air. 2. N2O emission affecting factors in MABR unit process Table 3 lists the main operational conditions and biofilm characteristics of MABR unit process at these plants, which were grouped into mainstream and sidestream applications. Among all these conditions, bulk NO2--N concentration is identified as the key factor affecting N2O emission for MABR technologies. This seems reasonable because nitrite affects the activities of all bacteria involved in N2O production (Kemmou, L. and Amanatidou, E., 2023). Figure 2a below show that a linear relationship (R2=0.7682) between the N2O EF and the bulk NO2-N concentration, which varied from 0 to around 120mgN/L. HigherNO2-N concentration results in higher N2O emissions probably because high NO2-N concentration might promote AOB denitritation and inhibit heterotrophs denitritation. Since bulk NO2--N concentration is relatively low in most MABR applications, a plot of the N2O EF at the bulk NO2-N concentration less than 4mgN/L is carried out (Figure 2b). Figure 2b shows that the impact of bulk NO2-N concentration become weaker when it becomes lower than 4mgN/L. 3. Impact of MABR unit process on downstream unit processes MABR unit process might have a positive impact on the N2O emission in the downstream aerated tanks due to process intensification by the MABR biofilm. The monitoring campaign at Site 1, which compares the N2O emissions between two identical lanes (Site 1a and 1b), indicated that the lane with MABR retrofitting (Site 1a) had much lower N2O EF. Further studies should be carried out to confirm and understand the impact of MABR unit process on the whole plant. Conclusions Data from eight full-scale MABR plants demonstrates, 1. MABR technologies reduce N2O emissions by approximately 2/3 compared with IPCC 2019 benchmark. 2. MABR exhaust is the main source of N2O emissions in MABR unit process, making it easy for gas-phase disposal to further reduce N2O emissions. 3. Bulk NO2-N concentration is the key affecting factor for N2O emissions in MABR technologies. However, once it becomes low, the impacts become weaker. 4. Counter diffusion in MABR biofilm plays a critical role in N2O mitigations.

This paper was presented at WEFTEC 2025, held September 27-October 1, 2025 in Chicago, Illinois.

Presentation time

14:00:00

14:30:00

Session time

13:30:00

15:00:00

SessionFull-Scale MABR Successes in Performance Optimization and N2O Control

Session locationMcCormick Place, Chicago, Illinois, USA

TopicLiquid Stream Treatment - Nutrient Removal and Recovery

Author(s)

LONG, ZEBO, Donnaz, Sylvain, Guo, Hui, Coutts, Daniel, Ireland, John, Peeters, Jeff, Hu, Neil

Author(s)Z. LONG1, S. Donnaz1, H. Guo1, D. Coutts1, J. Ireland, J. Peeters, N. Hu

Author affiliation(s)Veolia WTS USA Inc.1

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Oct 2025

DOI10.2175/193864718825160091

Volume / Issue

Content sourceWEFTEC

Word count15

MBR-DAS: Densification Improves MBR Performance at the City of Detroit Lakes

Abstract

Membrane bioreactors are often selected as secondary treatment technologies where small footprint and water reclamation are desired. Sludge densification by means of an external selector such as memDENSE with adequate biological configuration helps maximize sludge densification and quality, thereby improving biological performance. We investigated the resulting oxygen transfer efficiency and membrane performance of an MBR before and after upgrade with densification and recorded an improvement in aeration performance and permeability. Densified biomass has been shown to be more resilient and efficient than CAS in terms of overall settleability performance (SOR, SLR, sludge blanket level, RAS rate, RAS concentration; Donnaz et al, 2020; Roche et al, 2021). MBR mixed liquor quality can be enhanced by using hydrocyclones as gravimetric selectors, whereby MBR performance has been improved due to the removal in overflow of lighter flocs and colloidal fraction from the mixed liquor (Nogushi et al, 2018). External gravimetric selectors help regulate particles and granules size below 1 mm, which is believed to minimize pore blocking (Zhang & Jiang, 2018). MLSS was found to be the main parameter controlling the α-factor in MBRs, corroborating several studies which have established the negative effects of solids concentration on oxygen transfer (inter alia, Krampe and Krauth, 2003; Germain et al., 2007). However, although several investigators focused on the concentration and viscosity of the sludge (e.g., Cornel et al, 2003; Racault et al, 2011; Amaral et al, 2018), none included MBRs with densified sludge. We investigated here the effects of densification on oxygen transfer in a full-scale MBR. Densified biomass is expected to have beneficial impacts on MBR membrane filtration performance. Particle size distribution of densified biomass has shown to be 200 to 500 µm in majority, with 1,000 µm maximum (Roche et al, 2021). An external gravimetric selector such as memDENSE technology helps regulate particle and granule sizes above 100 µm and below 1 mm, which is believed to minimize pore blocking fouling observed with full granulation process by remaining below the 1 to 1.2 mm critical value range described by Zhang & Jiang (2018). A full-scale evaluation has been underway at the City of Detroit Lakes for over nine months to capture the benefits of biomass densification in a plant with challenging seasonal operation. Figure 1 shows the configuration of this MBR plant: anaerobic, anoxic and aerobic tanks in series (A2O configuration), with memDENSE incorporated as the external selection process on the RAS stream. Comprehensive characterization of mixed liquor properties and membrane fouling is being conducted before and during densification in this plant. The memDENSE system accumulates the large and dense biomass in the underflow while wasting the small and light biomass from the overflow (Fig. 1). In the first stage of intensification, the particle size and density increased in both underflow and overflow. Preliminary testing at full-scale at Detroit Lakes MBR facility showed early improvement of membrane filtration correlated to rapid drops of SVI, time-to-filter and superfine colloids (sfCOD) in the membrane tank mixed liquor (Donnaz et al, 2023). Time-to-filter is a direct test of filterability and membrane performance. Results of these tests show a very close correlation with settleability, meaning that improved settling and higher densification also improve filterability. With results similar to those of SVI, Figure 2 shows that poorly filterable sludge is wasted while a much more stable biomass is achieved with densification. This led to a significant improvement in the MBR performance. Preliminary results shown in Figure 3 indicate that with selective wasting and continuous flow densification, permeability is improved and TMP stabilized. Allowing an acclimation period of 4 months of memDENSE operation to achieve a new early steady state, oxygen transfer efficiency (OTE, % and αSOTE, %) was compared at the two process conditions. The OTE was measured using the ASCE method (ASCE, 2018). Since the clean water performance of the disc diffusers was available, the alpha factors could be calculated and compared. An increase in air flux after densification corresponds to increased oxygen uptake (Table 1). The efficiency (OTE and αSOTE) is higher, despite the increase in aeration, indicating an improvement in the quality of the biomass-wastewater matrix. In fact, the α-factor improved and is now reaching the design value of 0.50, despite the diffuser aging. In the presentation, we will discuss in detail the field results and their comparison with the design criteria.

MBR coupled with Densified Activated Sludge (memDENSE) tailors mixed liquor characteristics to optimise membrane & biological performance. memDENSE provides lower OPEX, the lowest and most stable TTF and TMP on the market. Preliminary study of biological sludge wasting (WAS) characteristics from the DAS system shows clear improvement in terms of thickening ability, dewaterability and biogas potential of the DAS sludge.

SpeakerShaw, Chris

Presentation time

13:30:00

14:00:00

Session time

13:30:00

15:00:00

SessionOptimization of MBR Technology

Session number210

Session locationRoom 343

TopicAdvanced Level, Facility Operations and Maintenance, Municipal Wastewater Treatment Design

Author(s)

Shaw, Chris, Donnaz, Sylvain, Bredeson, Rob, Danzl, Susan, Guo, Hui, Peeters, Jeff, Rosso, Diego

Author(s)C. Shaw1, S. Donnaz1, R. Bredeson2, S.R. Danzl4, H. Guo5, J.G. Peeters6, D. Rosso7

Author affiliation(s)1Veolia Water Technologies and Solutions, Ontario, 1, 2City of Detroit Lakes, MN, 3Veolia WTS, MT, 4, MN, 5Veolia WTS, Ontario, 6Veolia Water Technologies & Solutions, ON, 7University of California, Irvine, CA

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Oct 2024

DOI10.2175/193864718825159555

Volume / Issue

Content sourceWEFTEC

Word count12

What Particle Size Do You Need in Your Densified Systems

Abstract

[b]Introduction[/b] Densification of biomass has been demonstrated to be efficient in improving sludge settleability which determines the design and performance of secondary clarifiers (i.e., the hydraulic capacity) in wastewater treatment plants (WWTPs) (Roche et al., 2022). Densification processes by selectors contributes to a more stable biological treatment process, lower operation costs, and higher nutrient removal efficiency (Mielczarek et al., 2023; Li et al., 2019; Guo et al., 2010; Martins et al., 2024). Recently, integrated applications of memDENSE[sup]TM[/sup](membrane bioreactor with physical selectors) and zeeDENSE[sup]TM[/sup] (membrane aerated biofilm reactor bioreactor with physical selectors) have shown significant improvements in the membrane permeability/filterability and chemical saving (Donnaz et al., 2024, Astrand et al., 2024). In addition, membrane aerated biofilm reactor (MABR) as a densification process has been proven to enhance nitrification at lower solid retention times (SRTs) because of seeding effects (i.e., nitrifiers that grow in the MABR's biofilm are sloughed). There are many factors that govern the performance of the densification systems. Particle size distribution (PSD) has emerged as a key structural parameter since it relates to biomass morphology and affects physical/biological treatment performance. Although previous studies indicated growing recognition of PSD as an important factor in wastewater treatment performance, limited studies investigate specific particle size ranges to full-scale SVI trends and nitrification activity. Therefore, this study integrates 2 years full-scale operational data with 20+ detailed PSD profiles to identify: (1) which particle size range most strongly associates with SVI behaviour; (2) how the PSD affects AOB (ammonia oxidizing bacteria) and NOB (nitrite oxidizing bacteria) activity, respectively. Which range benefits AOB and NOB activity more; (3) what are the relationships between particle size ranges with EPS and zone settling velocity (on-going study); (4) which particle size range will be enriched from biofilm detachment and how it contributes to the nitrification rate. These works establish evidence-based PSD thresholds that can support practical process control and densification strategies. [b]Methods[/b] The comprehensive characterization of the sludge in this study involved regular monitoring and detailed analysis across various operational parameters. The mass fraction of particles larger than 212 µm (i.e., DI in this study) and SVI were measured 3 times per week on the site following standard protocols. Samples were collected during different operating stages (i.e., various DI) and conducted the PSD analysis using Master Sizer 2000. AOB and NOB activities for different size ranges were measured on sites with fresh sieved samples. The EPS and settling velocity analysis are currently ongoing, the results will be included in future presentations. A standardized protocol named clean water test was developed and implemented to quantify the seeding effects and study the baseline performance of a MABR system. During the operation, the MABR tank was completely cut off and drained entirely. After filling with a secondary clarifier, scouring and mixing were restarted. Samples were collected every 20 minutes both from the inlet and outlet of the MABR unit and conducted AOB and NOB activity tests. [b]Results[/b] The PSD analysis results were summarized in Figure 1. Size 100 to 200 µm ranging from 30%-50% (volume base) is the dominant fraction while the DI (i.e., mass fraction of particle >212 µm, measured by sieve tests on site) varies from 30% to 60% (Figure 1). Size 200 to 550 µm is the main part when DI is over 60%. Particles smaller than 100 µm decreased significantly from 40-50% to 8-20%. Particle size changes more significantly on memDENSE[sup]TM[/sup] site since this site receives industrial wastewater consistently. The correlation matrix between SVI with various particle size ranges and DI was presented in Figure 2. SVI shows a very strong positive correlation with the fraction of particles smaller than 50 µm (correlation coefficient:0.91). That indicates that a higher proportion of particles smaller than 50 µm results in higher SVI or worse settleability. The most negatively correlated parameters for SVI are the particle size 200 to 550 µm. This suggests future densification processes should control d90 in the range of 200 to 550 µm. The highest AOB activities were observed in the particles smaller than 63 µm while the highest NOB activities were found in particles ranging from 106 to 500 µm (Figure 3). Considering both microbial activities and mass fraction of particles in the mixed liquor, the size ranges from 106 to 550 µm contributed to more than 65% of N removal. A higher N removal rate was observed at the zeeDENSE[sup]TM[/sup] site compared to the memDENSE[sup]TM[/sup] site. This result can be explained by the seeding effects of the MABR reactor (detailed studies are ongoing). Similar results were observed during the seeding effects quantification tests. Although increased activities were observed both for AOB and NOB (Figure 4), smaller particles favor AOB activity while larger particles favor NOB activity. More detailed analysis will be conducted and presented during the technical sections.

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

Presentation time

08:30:00

09:00:00

Session time

8:30:00

10:00:00

SessionParticle Size Effects in Intensified BNR Systems

Session locationErnest N. Morial Convention Center

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

Author(s)

Guo, Hui, Fyfe, Sheila, Donnaz, Sylvain

Author(s)H. Guo1, S. Fyfe1, S. Donnaz3

Author affiliation(s)Veolia, 1Veolia, 1

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Sep 2026

DOI10.2175/193864718825160552

Volume / Issue

Content sourceWEFTEC

Word count11

Base text size -

Guo, Hui

Site admin

About Guo, Hui

My Recent Activity