Alternative MBR configurations − anaerobic MBRs and MABRs
The vast majority of the membrane bioreactors installed across the world are the classical biomass separation type, where the membrane is used to separate the clarified water from the biological solids from an aerobic biological treatment process. However, there are exist a few possible technical variations:
The application of membrane separation of the biomass from an anaerobic process, as opposed to an aerobic process, was first conceived and implemented in the late 1980s as the 'ADUF' process (Anaerobic Digestion − Ultrafiltration), configured as a sidestream MBR (hence 'AnsMBR').
More recently, the immersed configuration (AniMBR) has been implemented (see Case study: Ken's Foods).
Although the AniMBR technology has been subject to extensive research (see Feature: Immersed anaerobic MBRs: are they viable?), the implementation of AnMBRs has been limited to a few industrial effluent applications where the high organic carbon concentrations make the process energetically favourable. This appears to be partly because of the highly fouling nature of the anaerobic biomass, reducing the membrane permeability and thus the permeation energy demand and/or membrane area requirement.
The anaerobically-treated effluent also normally demands further treatment for removing the nutrient content, since anaerobic processes provide little or no nutrient removal. The key advantage offered by the membrane separation of a high-clarity water is thus lost if further biological processing for removing the nutrients is required.
Available information suggests that AnsMBRs can provide a COD rejection of 99% or more, the % removal increasing with increasing feedwater concentration, and achieve a flux of 15−30 LMH for a range of food effluent applications.
Operating conditions (crossflow velocity, CFV, and transmembrane pressure, TMP) appear to be similar to those employed for an aerobic sMBR, with a reduction in the CFV producing a corresponding reduction in the sustainable flux. This being the case, the value offered by anaerobic as opposed to aerobic treatment by a conventional sMBR is determined by the balance of (from the perspective of the anaerobic option):
- the OPEX benefit of the methane generated, which is then proportional to the difference in the feed and permeate COD concentration
- the OPEX benefit of the reduced process aeration (assuming all other aspects of the anaerobic and aerobic biological process OPEX to be similar)
- the OPEX benefit of the reduced sludge production
- the OPEX penalty of the increased specific energy demand for the membrane filtration (which is proportional to the flux)
- the CAPEX penalty associated with the larger membrane area demanded by the lower flux, and
- the overall cost penalty of supplementary downstream nutrient and residual COD removal, if required.
Since flux does not appear to be a function of loading, the anaerobic MBR option − as with the classical treatment − becomes more viable at higher loadings. This arises from a combination of the calorific value (CV) of the methane generated (1) and the reduction in process aeration (2), both of which are roughly linearly related to the COD (as is the proportional reduction in sludge (3)). The OPEX penalty (4), for a pumped sMBR, roughly equates to the permeability, whereas the CAPEX penalty is inversely proportional to the flux.
It is because of the significant OPEX penalty that there has been recent interest in the immersed configuration (the aniMBR), which demands a much reduced energy for permeation and for which scouring can potentially be provided by the generated biogas. Pilot-scale studies of this configuration, along with data from a full-scale installation, suggest that aniMBR fluxes are generally in the range of 4−10 LMH depending on the feedwater quality. There is also some indication from iHF studies that backflushing may significantly increase the sustainable flux.
Membrane aerated biofilm reactors (MABRs)
In an MABR, the membrane (usually a hollow fibre) is used as an alternative aerator, rather than for biomass separation. As such, it does not provide the highly-clarified effluent of a classical MBR: an additional membrane separation stage would be required for this.
The term ‘MABR’ was first introduced in the 1990s by Mike Semmens of the University of Minnesota, and was originally based on a pure oxygen-fed technology. The MABR allows almost all the oxygen introduced to the bioreactor via the membrane to be utilised by the micro-organisms since it is introduced in molecular (or ‘bubbleless’) form. Because a biofilm is formed on the membrane itself, the oxygen is delivered directly into the biomass. The usual mass transfer limitations of a conventional fine bubble diffuser aerator (FBDA), limiting the standard oxygen transfer efficiency (SOTE) to somewhere between 10 and 40%, therefore no longer apply.
While the membrane provides greatly increased aeration efficiency, and commensurately reduced aeration energy costs, the process is hampered by the high cost of the membrane, compared to a conventional FBDA, and the loss of mixing provided by the air bubbles introduced by the FBDA. Also, as with conventional HF membranes, the membranes can clog as a result of the biofilm growth on the membrane surface, adding to the process operational complexity.