Membranes Vs Concrete: Defining the value and limitations of hybrid MABR retrofits

The capability of membrane aerated biofilm reactors (MABRs) to be retrofit into existing basins provides a unique value proposition for legacy WRRFs with tightening regulations or capacity limitations. For plants seeking more nitrogen removal, MABRs allow for development of long-SRT biofilms which can both nitrify and denitrify, decoupling nitrification capacity from the aerobic SRT of the suspended biomass. This adds nitrification capacity without altering the plant's suspended biomass inventory or increasing loading on secondary clarifiers. Placement of MABRs in unaerated zones also allows this technology to be coupled with low-DO aeration controls and simultaneous nitrification-denitrification configurations that don't require mixed liquor recycles (IMLR). In combination, these intensified process configurations allow for low-energy and carbon-efficient treatment in addition to retrofittable capacity increases.

These benefits, described in the literature mainly in general terms, are now being realized with implementation of full-scale hybrid MABR retrofits. Using a combination of historical data and process modeling, this paper quantifies the benefits and identifies design best-practices for spiral-configured MABR retrofits, both for a real installation in Israel, and more generally, for a typical North American plant located in the temperate climate zone.

Roughly two years of water quality and operational data were analyzed from a hybrid, spiral MABR retrofit in the anoxic zone of the Maayan Tzvi WRRF, an A2O plant operated at low-DO setpoints (DO<0.5 mg/L) and without IMLR. Performance results, as measured by the quantity of ammonia oxidized by MABR versus suspended growth at the plant, demonstrate that MABR met or exceeded the expectations established by earlier piloting, though the impact of the MABR was somewhat variable over time (Fig. 1). Influent, effluent and operational data were used to develop SUMO models of this plant in both MABR and 'conventional' (no MABR) configurations. Using real-world data within a dynamic process model allowed for investigation of factors contributing to biofilm nitrification rates, which can serve as a measure of process efficiency and operational performance.

SUMO analysis found that, for a membrane configuration designed to nitrify 30% of influent NH4 (as is the vendor's standard,) performance improved with decreased organics loading and increased NH4 loading, but only up to a maximum of 10 g N/d*m2 (Fig. 2, 3). At higher NH4 loadings and concentrations (such as may occur in dry regions with established water conservation programs), modelled system performance diminished during diurnal periods of increasing nitrogen loading (roughly 11:00-19:00 in this dataset). This finding suggests that for certain periods of the day in these geographies, nitrification may not be substrate-limited as generally assumed, but rather limited by some other factor. Oxygen limitation is one potential contributor, as nitrification rates rebounded when modelled rates of oxygen transfer through the membrane were set well-above the data-derived values. Previous findings of similar inhibitory effects at full scale identified low ORP or high sulfur concentrations as likely factors (Uri-Carreño et al. 2023), however neither were reflected in this study's modeling results. While additional high-resolution sampling of full-scale retrofits would help to confirm these findings, they nonetheless suggest that designers consider using IMLR, equalization, increased membrane area, or other techniques for maintaining instantaneous MABR substrate loadings below a threshold value of 10 g N/d*m2 throughout the diurnal cycle to avoid inhibitory effects on performance when using these commercially available membranes.

When considering the benefits of MABR to designers, this study found that in temperate regions (design water temperature of 15°C), addition of MABRs tasked with removal of up to 30% of influent ammonia reduces the required aerobic SRT of the suspended biomass system by two full days, while maintaining a 2x nitrification safety factor (Fig. 4). This represents a 25% reduction in the required secondary clarifier capacity (achieved through reduction in MLSS), or a 14% reduction in the aerobic basin concrete volume (Fig. 5). In both cases, reducing the need to construct additional basin volume or clarifiers results in GHG emission savings above and beyond the savings in aeration energy. In warm (>30°C) climates, by contrast, this benefit is lost. By combining MABR with low-DO operation, a total energy savings of 30% can be achieved regardless of climate (Fig. 6), while also producing a small (roughly 3mg/L) improvement in effluent total nitrogen.

This study combines modeling with fullscale data from a successful hybrid MABR retrofit to identify design factors which influence the performance of such systems under various temperature and loading regimes. This information will be useful to designers wishing to maximize the value derived from MABR retrofits of existing assets. By translating dynamic performance metrics into reductions in minimum aerobic SRT, this study also helps to define the value proposition of MABR capital investments as compared to construction of additional tankage. As understanding of this value proposition improves, MABR is likely to become a more widely adopted technology for retrofit of capacity-constrained BNR plants.