Shaw, Chris

Chris Shaw is a speaker at WEFTEC 2026.

Titles from this speaker

Application of RIS-Based Soft Sensor for Comprehensive DAS-MBR Resilience Insights

Abstract

Recently, integrating Densified Activated Sludge (DAS) with Membrane Bioreactors (MBRs) has shown promise for improving mixed liquor filterability and reducing membrane fouling [1]. MBRs improve on conventional activated sludge (CAS) systems by using membrane filtration for treatment and solid–liquid separation, but membrane fouling remains a challenge, reducing permeability and increasing energy use and costs [2-3]. Converting CAS mixed liquor to a DAS system forms and retains denser biological flocs, which selectively remove pin floc and filamentous bacteria associated with poor settling and filterability. These denser flocs also reduce the effects of EPS, SMP, and colloids, resulting in lower fouling potential and energy consumption [4-5]. Technologies like memDENSE hydrocyclones act as physical selectors, retaining denser granules while removing lighter sludge [5]. DAS implementation can boost process capacity and resilience to shock loads and toxins, offering a practical solution for modern wastewater treatment [1, 6]. Membrane fouling can be classified by cleaning strategy into reversible (removable by air scouring), residual reversible (requiring backwashing), irreversible (requiring chemical cleaning), and irremovable (permanent) [7]. Fouling is often quantified with the resistance-in-series (RIS) model, which uses Darcy's law to sum membrane and fouling resistances using transmembrane pressure (TMP) measurements [8-9]. This study introduces a new adaptation of the RIS model as a soft sensor to provide deeper, data-driven insights into MBR resilience, particularly with DAS integration. This approach enables a more thorough evaluation of membrane resistance and system robustness, improving comparisons between CAS-MBR and DAS-MBR systems at full scale. A full-scale comparison of CAS-MBR and DAS-MBR systems was performed at the Detroit Lakes, Minnesota WWTP. CAS-MBR was monitored for 28 months, then DAS-MBR for another 28 months. The facility has a design capacity of 2.2 MGD and operates in an A2O configuration using ZeeWeed 500D membranes. Waste activated sludge from the return activated sludge line is processed through memDENSE hydrocyclones, which retain dense biomass in the underflow for recombination with the plant influent, while lighter biomass is wasted from the overflow. The DAS system was implemented in April 2023. Each membrane filtration cycle lasts 12 minutes, with 1 minute dedicated to cleaning via LEAP aeration air scouring. If the TMP exceeds the threshold, the air scouring is replaced by a manually triggered back pulse. The total resistance to filtration (R[sub]tot[/sub]) (m⁻¹) is calculated using Eq.1. where A[sub]m[/sub] is the membrane surface area (m²), F is the permeate flow rate (m³/h), µ is the permeate viscosity (assumed constant and equal to that of water at 20°C, in Pa·s), T is the temperature (°C), and C[sub]t[/sub] is the temperature correction coefficient (1.033 if T<20°C and 1.025 if T>20°C). In this study, the RIS model, which breaks down R[sub]tot[/sub] into its components, is refined as Eq.2. where R[sub]int[/sub] is the clean membrane intrinsic resistance, R[sub]irr[/sub] is the irremovable resistance, R[sub]hyd[/sub] is the hydraulic resistance (including both reversible and residual reversible fouling), R[sub]mc[/sub] is the resistance removable by chemically enhanced backwashing (CEB), and R[sub]rc[/sub] is the resistance removable by chemical recovery cleans. R[sub]int[/sub] cannot be directly measured. For the soft sensor implementation, R[sub]hyd[/sub] is calculated as the difference between total resistance before and after a hydraulic clean (air scour or backwash). R[sub]mc[/sub] is calculated as the difference between the 6-hour average total resistance after a hydraulic clean, before and after each CEB. Long-term RIS model analysis at Detroit Lakes WWTP found that switching from CAS-MBR to DAS-MBR reduced backwashing frequency and improved resilience. Figure 1 shows monthly averages for total, hydraulic, and CEB resistance. Seasonal industrial discharges from February to June increased influent orthophosphate, total COD, and COD<1.5µm, leading to higher resistance values. After memDENSE installation, these seasonal effects on membrane resistance decreased, resulting in a more stable system. Although resistance rose slightly from August to October 2024 due to high festival loads, the system remained resilient with only minor increases in resistance. Figure 2 shows the monthly distribution of total membrane resistance, averaged across all four MBR trains, along with the number of backwashes performed each month. Notably, from June 2023-two months after DAS implementation-no backwashing was needed, except during the system upset from August to October 2024, when a few backwashes were performed to stabilize operations. After implementing DAS, CEB resistance for both sodium hypochlorite and citric acid remained stable, allowing for a 50% reduction in cleaning frequency from May 2025 without compromising system performance. Figure 3 shows that annual chemical use for backwashing, normalized per 1000 m³ of permeate, decreased compared to previous years. In conclusion, this study applied an adjusted RIS model as a soft sensor to assess membrane resistance in full-scale CAS-MBR and DAS-MBR systems. As Figure 4 shows, after memDENSE installation in 2024–2025, both membrane resistance and backwashing frequency decreased by about 70%.

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

SpeakerShaw, Chris

Presentation time

09:00:00

09:30:00

Session time

08:30:00

10:00:00

SessionSmart MBR

Session locationErnest N. Morial Convention Center

TopicCircular Water Economy, Facility Operations and Maintenance, Municipal Wastewater Treatment Design, Nutrients, Research and Innovation

Author(s)

Mohebali, Sanaz, Shaw, Chris, Gagnon, Jean, Alsayed, Ahmed, Donnaz, Sylvain

Author(s)S. Mohebali1, C. Shaw1, J. Gagnon, A. AlSayed2, S. Donnaz5

Author affiliation(s)Veolia, 1Veolia Water Canada, 1Western University, 2

SourceProceedings of the Water Environment Federation

Document typeConference Paper

PublisherWater Environment Federation

Print publication date Sep 2026

DOI10.2175/193864718825160440

Volume / Issue

Content sourceWEFTEC

Word count11

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

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

Base text size -

Shaw, Chris

Site admin