107
Effect of membrane type on MABR performance: silicone
108
Figure 1. Cross sectional view of micromembrane cord (left) and cross sectional view of silicone membrane (right)
Silicone hollow-fiber membranes and micromembrane cords may differ in terms of mass transfer coefficient Km, areas of membrane gas-transfer and attachment, ability to retain biomass, and type of biofilm development. Higher Km values are desirable, as they allow higher contaminant removal fluxes. Biofilms with higher densities generally have higher removal fluxes. Also, rough biofilms can result in lower fluxes for counter-diffusional biofilms. Higher gas transferring area and attachment area can be beneficial. The membrane surface chemistry and roughness affect attachment.
Research is needed to explore these differences and their practical implications.
In this research, we explored the nitrification fluxes in silicone membranes and micromembrane cords. The membranes were similar, but not necessarily identical, to those currently used by Suez and Oxymem. We used single-membrane flow cells to determine the Km values and explore biofilm development, and reactors with membranes bundles to assess removal fluxes.
Materials and methods
Single-membrane flow cells and membrane bundle tube reactors in the lab were used and experiments were run concurrently with either silicone membranes or micromembranes provided by Suez. The bundle reactors had N2 bubbling, promoting biofilm detachment. The Km values were calculated accrording to Eq. 1, the biofilm roughness was calculated using Eq. 2, and the nitrification flux was calculated using Eq. 3.
𝐾𝑚 =
𝐷𝑂2𝑑𝐶𝑂2,𝑙
𝑑𝑟 |
𝑟=𝑅𝑚 𝐶𝑂2,𝑚(𝑔)
𝐻 −𝐶𝑂2,𝑚(𝑙) Eq. 1
𝑅𝑎′ = 1
𝑁∑ (|𝐿𝑓,𝑖−𝐿̅̅̅̅|𝑓
𝐿𝑓
̅̅̅̅ )
𝑁1 Eq. 2
𝐽𝑁𝐻4 =𝑉𝑟[(𝑁𝐻4∆𝑡 𝐴(𝑡)−𝑁𝐻4(𝑡1)]
𝑚 Eq. 3
where Km is the membrane mass transfer coefficient (m/d), CO2,m(g) and CO2,m(l) are the O2
concentrations (mg/L) on the gas side (g) and the liquid sides (l), and His Henry coefficient for O2, 𝑅𝑎′
is the biofilm roughness coefficient, N is the number of measurements, 𝐿𝑓,𝑖 is the biofilm thickness at point i, and 𝐿̅̅̅𝑓 is the average biofilm thickness, Vr is the reactor volume (m3), NH4(t)-NH4(t1) is the change in NH4+ concentration(gN/m3), over the time interval ∆t (d), and Am is the membrane area (m2) Results and discussion
Dissolved oxygen profiles were taken with microsensors perpendicular to the membrane surface for the single silicone membrane and micromembrane cord to calculate their Km values (Figure 2).
Equation 4 was used to calculate the Km using the slopes of the measured DO profiles. The average Km value for the silicone membranes was 3x10-5 m/s (2.6 m/d) and for the micromembrane cords, it
109 was 5.6x10-6 m/s (0.4 m/d). The higher Km value for the silicone membrane indicates it can transfer oxygen at higher rates, for a give concentration differential.
Figure 2. Typical DO profiles for membranes without biofilm for the (a) micromembrane cord and (b) silicone membrane.
Images of biofilm development were taken in the single membrane flow cells (Figure 3a). Biofilms on the silicone membranes grew thicker and with a lower density (Figure 3b), possibly due to higher oxygen concentrations resulting from the higher Km. However, in experiments with bundles the micromembranes displayed thicker biofilms, possibly due to better attachment to the rougher surface under the more aggressive sher conditions in the bundle reactors. Excessive biofilm thickness in MABRs can reduce the removal rates 7,11–13 so choice of membrane material may affect biofilm thickness. The biofilms growing on the micromembrane cords were rougher than the silicone ones (Figure 3c), other than on the third day, when bubbles were observed on the silicone membrane, skewing the biofilm thickness measurements (Figure 3a).
Figure 3. (a) Images of the development of heterotrophic biofilms on micromembrane cords (top) and silicone membranes (bottom) over nine days. Heterotrophic biofilm (b) thickness and (c) roughness over 9 days
For the membrane bundles, the silicone membrane had significant variability in biofilm thickness along the membrane length. The micromembrane biofilms were more uniform, but with a lower average thickness. The turbulence and membrane collisions caused by N2 sparging appeared to cause greater detachment in the silicone MABR, which may have been caused by potentially higher eukaryotic predation rates due to higher DO values, or the smoother membrane surface.
0 1000 2000 3000 4000 5000
0 5 10
Biofilm Thickness (µm)
Time (Days) Micromembrane Silicone
0 0.2 0.4 0.6 0.8
0 5 10
Roughness
Time (Days)
110 Batch tests were carried out to characterize the ammonium oxidation fluxes at different O2
pressures and bulk ammonium concentrations (Figure 4).
Figure 4: Model predicted fluxes (continuous lines) and experimental results (dots) for (a) the silicone MABR (Km = 2.6 m/d, biofilm thickness 250 ± 60 µm) (b) the micromembrane MABR (Km = 0.4 m/d, biofilm thickness 330 ± 140 µm) at different bulk NH4+ concentrations and intramembrane relative pressures of 20, 40, and 60 kPa.
The experimental fluxes for two membranes are not directly comparable, as the biofilm thicknesses were different. However, the trends were similar: fluxes increased with increasing bulk NH4+, for a given air supply pressure, until approaching a maximum flux, when the biofilm becomes saturated with ammonium. Higher air pressures led to higher fluxes, especially at higher bulk NH4+
concentrations, because under these conditions the biofilm is O2 limited. The silicone MAB had higher maximum fluxes, because of the higher Km for the silicone membrane: 2.6 m/d, which allowed for higher oxygen concentrations and therefore larger ammonium fluxes. Also, despite the thicker biofilm, the maximum fluxes for the micromembrane cords were reached at lower bulk NH4+ concentrations.
This probably was due to the lower Km for the micromembranes, which results in lower DO concentrations at membrane surface for a given pressure.
In conclusion, the silicone membranes had a higher Km, allowing higher fluxes for a given air pressure. The micromembrane developed thinner biofilms and were better able to retain biomass under the turbulent bundle conditions. The two membranes had similar fluxes at low bulk ammonium concentrations (i.e., below 5 mgN/L). But silicone provided substantially higher fluxes at bulk ammonium concentrations above 10 mgN/L. Note these results may not translate directly to observed behavior in commercial systems. These may be impacted by oxygen gradients along the membrane, cassette configuration, mixing conditions, and other factors not assessed in our studies.
References
1 T. Ahmed, M. J. Semmens and M. A. Voss, Oxygen transfer characteristics of hollow-fiber, composite membranes, Advances in Environmental Research, 2004, 8, 637–646.
2 M. Aybar, G. Pizarro, J. P. Boltz, L. Downing and R. Nerenberg, Energy-efficient wastewater treatment via the air-based, hybrid membrane biofilm reactor (hybrid MfBR), Water Sci Technol, 2014, 69, 1735–1741.
3 M. J. Semmens, Membrane Technology: Pilot Studies of Membrane-Aerated Bioreactors, Water Environment Research Federation.
4 E. Syron, M. J. Semmens and E. Casey, Performance analysis of a pilot-scale membrane aerated biofilm reactor for the treatment of landfill leachate, Chemical Engineering Journal, 2015, 273, 120–129.
5 J. Wu and Y. Zhang, Evaluation of the impact of organic material on the anaerobic methane and ammonium removal in a membrane aerated biofilm reactor (MABR) based on the multispecies biofilm modeling, Environ Sci Pollut Res, 2017, 24, 1677–1685.
6 K. J. Martin and R. Nerenberg, The membrane biofilm reactor (MBfR) for water and wastewater treatment:
Principles, applications, and recent developments, Bioresource Technology, 2012, 122, 83–94.
111
7 M. J. Semmens, K. Dahm, J. Shanahan and A. Christianson, COD and nitrogen removal by biofilms growing on gas permeable membranes, Water Research, 2003, 37, 4343–4350.
8 E. Syron and E. Casey, Membrane-Aerated Biofilms for High Rate Biotreatment: Performance Appraisal, Engineering Principles, Scale-up, and Development Requirements, Environ. Sci. Technol., 2008, 42, 1833–1844.
9 P. Perez-Calleja, M. Aybar, C. Picioreanu, A. L. Esteban-Garcia, K. J. Martin and R. Nerenberg, Periodic venting of MABR lumen allows high removal rates and high gas-transfer efficiencies, Water Research, 2017, 121, 349–360.
10 R. Nerenberg, The membrane-biofilm reactor (MBfR) as a counter-diffusional biofilm process, Current Opinion in Biotechnology, 2016, 38, 131–136.
11 E. Casey, B. Glennon and G. Hamer, Biofilm development in a membrane-aerated biofilm reactor: effect of intra-membrane oxygen pressure on performance, Bioprocess Engineering, 2000, 23, 457–465.
12 O. Debus and O. Wanner, Degradation of Xylene by a Biofilm Growing on a Gas-Permeable Membrane, Water Science and Technology, 1992, 26, 607–616.
13 R. Wang, F. Xiao, Y. Wang and Z. Lewandowski, Determining the optimal transmembrane gas pressure for nitrification in membrane-aerated biofilm reactors based on oxygen profile analysis, Appl Microbiol Biotechnol, 2016, 100, 7699–7711.
Presenting Author
Emily Clements PhD Student
University of Notre Dame
Is the presenting author an IWA Young Water Professional? Yes Bio: Emily is a PhD student at the University of Notre Dame, studying the effect of temperature on biofilm systems, the factors influencing nitrification fluxes in MABRs, and modelling water age in premise plumbing systems.
112
Membrane Aerated
Biofilm Reactor Session
Poster presentations
113