Bringing It All Together: Designing a Densified AO/SND Process for Efficient Biological Nutrient Removal

ABSTRACT: This study presents the design, challenges, and process control strategy for low DO BNR facility with DAS in cold climate using ABAC. The advantages of this design include simplification of a 4-stage Bardenpho process to an A/O configuration with EBPR, reducing operational complexity and air demand while still maintaining effective nitrogen removal. This is the first application of low DO process in cold climate with daily effluent ammonia limit using DAS and ABAC. The full-scale construction began in the third quarter of 2023 and will be completed by May 2025. Start-up activities and process changes to low DO will begin in January 2025.

FACILITY BACKGROUND: The secondary process at Boulder WRRF includes anoxic/aerobic/post-anoxic zones followed by Solids Contact Tanks (SCTs). BWRRF is currently under construction to upgrade the secondary process with DAS, low DO A/O configuration to promote SND with inDENSE™ to meet regulation-driven targets of 0.75 mg/L-P and 9.0 mg/L-N.

PROCESS DESIGN: The process design includes upgrades for A/O SND process with ABAC to meet daily effluent ammonia and TIN limits. Installation of supplemental mixing, replacing all remaining ceramic disc membranes with membrane discs, and incorporating inDENSE™ are also included in the design.

Basis of Design: The basis of design for the selector is presented in Table 1. The selector has an important role in densification, filaments control, and EBPR. Key design considerations for feast and famine ratio, F/M ratio, HRT, and SRT are noted in Table 1. Zone 4 is designed as a swing zone to augment anaerobic volume if needed. The F/M and feast and famine ratios selected for design for this project is consistent with other similar applications (Strum, 2020). Feast/famine ratio for this application is at 20% (for three selector zones). The criterion used for determination of feast/famine was provided in Sun et al. (2022).

Mixing System: Dynamic modeling was used to develop minimum airflows to maintain ML suspension. Typically, 0.05 to 0.09 SCFM/SF is used for facilities with primary clarification. Design guidelines for mixing requirements for DAS is lacking; however, it is assumed that DAS biomass is similar to CAS. Table 2 compares the minimum airflow with model results. Given the concerns over potentially overaerating, and the importance of maintaining a suspended biomass during periods of low load, a supplemental large bubble mixing system was included in Zones 4 and 8, and the SCTs (Figure 2). The large bubble mixing system compressors were retrofitted into an existing, unused blower room and include desiccant dryers. The dried airstream is then delivered by exposed piping to the aeration basins and SCT.

Densification Technology Design: Physical selection is important for accumulation of granules (Roche et al., 2022). For this project, hydrocyclones will be used in an effort to improve settleability and low DO process performance. The hydrocyclones have the flexibility to pull either ML from the ML channel or RAS and includes the ability to remotely add or remove hydrocyclones from operation. These design features provide additional flexibility to control wasting. The design criteria for the hydrocylones are presented in Table 3. Figure 3 shows the full range of wasting capacity. Biological selection which is exposing specific bacteria to alternating cycles of feast and famine is critical (WRF, 2024). The alternating A/O condition is intended to promote the growth of PAO and retention by hydrocyclone to increase granule density and settling velocity.

Aeration Blowers: The process air blowers provide oxygen to BNR process and a sidestream Post-Aerobic Digester. The existing blowers were oversized, difficult to turndown, and reaching their end of useful life. Two of the existing blowers will be replaced with dual-core turbocompressors to improve flexibility and turndown. Additionally, a common header was bisected to allow for blowers to be dedicated to a single aerobic process, allowing for additional energy savings due to lower header pressures.

Process Modeling: A calibrated dynamic process model was developed in BioWin, SUMO, and SIMBA#. Table 4 compares the process performance in BioWin and SUMO. Details of process modeling is presented in another paper. Modeling was conducted with default and adjusted kinetics. For kinetics adjustments, KDO of AOB and NOB of 0.4 and 0.2 mg O2/L were used, respectively. The model with default kinetics provided slightly better nitrogen removal performance. Sensitivity analysis was also conducted in BioWin by including hydrocyclones with specified removal for PAO of 1.5 to study the impact on EBPR. Modeling was conducted with ABAC feedback strategy in BioWin and feedback and feedforward (step gain) in SIMBA#. Both models showed that the proposed design is able to consistently meet effluent ammonia of <1 mg/L. The advantages and disadvantages of each control strategy are discussed.

SUMMARY: With this upgrade, BWRRF will convert a complex 4-stage Bardenpho process to an A/O configuration, and this shift not only simplifies the treatment process but also eliminates the need for internal ML recycle. Advantages are reducing operational complexity and air demand while still maintaining effective nitrogen removal. This is the first application of low DO in cold climate with daily effluent ammonia limit with DAS and ABAC.