Process Intensification At The WNWRP: Evaluating MABR With Piloting and Process Modelling

Background
The Sanitation Districts are evaluating the potential of the MABR process to increase the biological treatment capacity of the Whittier Narrows Water Reclamation Plant (WNWRP). The WNWRP has a design capacity of 15 million gallons per day (MGD) and employs the Modified Ludzack-Ettinger (MLE) process to achieve nitrogen removal (Figure 1). The plant currently treats approximately 9 - 10 MGD of wastewater. National Pollutant Discharge Elimination System (NPDES) permit requirements for nitrogen, activated sludge process general design criteria, and typical activated sludge process operating conditions are shown in Tables 1, 2, and 3, respectively. WNWRP meets the NPDES permit requirements at current flows and loads but may struggle to meet the limits at the design capacity of 15 MGD, particularly with respect to nitrate + nitrite.
For this research project, the Sanitation Districts evaluated MABR using the ZeeLung product manufactured by SUEZ, described in Figures 2 - 5. Potential benefits of the MABR process include: (1) the potential to increase capacity within existing tanks while meeting NPDES limits for nitrogen; (2) the potential to increase capacity without increasing solids loading on the secondary clarifiers; (3) the potential to reduce aeration energy requirements due to efficient transfer of oxygen through gas-transfer media.
Objectives
The MABR process is being evaluated through pilot testing, subsequent development of a calibrated process model in BioWin process simulator, and planning level cost estimation for full-scale implementation. The specific objectives for pilot testing include: (1) evaluating the removal of organics, nitrogen and suspended solids over a range of flowrates representing current and potential future operating conditions; (2) evaluating key MABR performance indicators including oxygen transfer rate (OTR), oxygen transfer efficiency (OTE), and nitrification rate (NR); and (3) collecting information necessary to develop a calibrated process model for plant design and planning level cost estimate.
Methodology Pilot System
Based on preliminary BioWin modeling projections by SUEZ (Figure 6), a pilot system was designed and constructed to simulate WNWRP operation with an MABR zone (Figures 7 and 8, Table 4). The pilot system included an anaerobic selector (3%), MABR zone (31%), aerobic zone (66%), and secondary clarifier. The MABR zone was equipped with a 3 module ZeeLung cassette provided by SUEZ (Figure 9). The resulting media packing density, 2.53 ft2/gallon, is similar to that assumed for the full-scale preliminary modeling projections, 2.73 ft2/gallon. The pilot system was equipped with sensors to continuously monitor ammonia in the MABR zone and O2 in MABR exhaust gas.
MABR Pilot System Operation and Performance Monitoring
The pilot system was operated in multiple phases (Table 5) with the ultimate goal being to maximize flowrate through the system while meeting permit requirements for nitrogen. All five phases of pilot operation are now complete. MABR performance parameters monitored include OTR, OTE, NR. Overall pilot performance was also being monitored by collection of operations and water quality data (Tables 6 and 7). Process Modelling A BioWin model of the WNWRP was created by the Sanitation Districts and calibrated to current plant flows, loading, and performance (Figure 10). The model was extended to 15 MGD to show predicted future performance. After the release of BioWin 6.2 containing an MABR specific element with a 1-D biofilm model, the WNWRP model was updated to include MABR in the anoxic zone of the plant (Figure 11). The MABR element in the BioWin model was calibrated using SUEZ developed calibration parameters (Table 8) and fine-tuned using operating data for OTR and NR collected during the pilot trial.
Results Pilot System
Phase 1 of pilot operation (Table 9) included start-up and biofilm development. OTE and OTR results are shown in Figures 12 and 13. The results indicate that an active biofilm was completely developed in approximately one month as reported in previous studies (Cote et al., 2015; Kunetz et al., 2016).
Phases 2 through 5 simulated various WNWRP operating conditions (~ 3 - 5 MGD/reactor, Tables 10 to 15) with the addition of a fully functioning MABR zone. OTR results are shown in Figure 14 and average results for OTR, OTE, and NR are shown in Table 16, along with results from previous studies for comparison. The potential to reduce aeration energy requirements is illustrated by comparing the OTE achievable by the MABR process to the OTE observed under field conditions for fine bubble diffusers at Sanitation Districts' facilities, which is typically approximately 10 - 20% depending on fouling condition (Wong et al., 2018). Previous studies have demonstrated that the primary factor impacting MABR performance is ammonia load (Cote et al., 2015; Houweling et al., 2017). This has been demonstrated by observing the diurnal response of OTR, NR, or OTE to primary effluent ammonia load variation (Figures 15 to 17). Results of nitrogen analyses conducted during Phases 2, 4A, and 5B are shown in Table 17. Average effluent ammonia and nitrate + nitrite concentrations were below permit requirements.
Modelling The process models of the WNWRP were run at both steady state and dynamic conditions, with steady state operating conditions summarized in Table 18 for both models. The current plant model was extended to run at 15 MGD and showed the plant would be unable to meet the NPDES permit for nitrate + nitrite limits under steady state or dynamic conditions, as shown in Figures 18 and 19 and Table 19. The MABR model showed that the plant can meet the effluent requirements, under both steady state and dynamic conditions, as shown in Figures 20 and 21 and Table 19. The model predicted effluent NOx-N of 8 mg/L was higher than the NOx-N achieved during piloting and refinement of the model will equate the performance between model and pilot. The model response of OTR and NR to ammonia concentration is captured in Figures 22 and 23, showing the model OTR and NR results are close to the results obtained during the pilot study. The improved performance with the MABR model can be attributed to: a) removal of ~28% of the influent ammonia and seeding effect of the biofilm on the bulk MLSS (Houweling et al, 2018); and b) simultaneous nitrification and denitrification in the MABR zone enabling increased denitrification capacity. The modelling results also show that the MABR process will use 15% less energy for aeration than the current plant at its' design capacity of 15 MGD. The pilot study and modelling exercise demonstrate that MABR is a viable technology to upgrade the capacity of the WNWRP while meeting the NPDES permit limits and saving energy. The results will be used as the basis for a planning level cost estimate, which be used in the planning of the potential MABR implementation at WNWRP.