Sustaining membrane permeability during unsteady-state operation of anaerobic membrane bioreactors for municipal wastewater treatment following peak-flow
Introduction
Anaerobic membrane bioreactors (AnMBRs) are a promising alternative to conventional aerobic biotechnology for municipal wastewater treatment, as the combination of organic degradation without the demand for aeration, coupled with energy recovery from biogas production, offers the potential to realise energy neutral wastewater treatment [1]. The key challenges limiting full-scale application of AnMBR for municipal wastewater treatment, are the membrane investment cost and energy demand associated with membrane fouling control [2]. Numerous previous studies have focussed on sustaining membrane operation through application of various hydrodynamic conditions (e.g. gas sparging rate, physical cleaning frequency and duration). In each of these studies, a steady-state influent flow rate is assumed, with the membrane fixed at constant flux [1], [3]. However, at full-scale, MBR must be designed to manage diurnal peaks and storm water flows [4]. Installation of equalisation tanks can serve to ameliorate peak flow and improve flow regulation [5]. Nevertheless, in a survey of 17 full-scale municipal aerobic MBR plants in Europe [6], half were reported to have peak ratios (peak flow to average flow) between two and three, due to the diurnal flow pattern and connection to combined sewer systems. The membrane must therefore be designed to cope with an increased flow without incurring substantial long-term fouling. This can be facilitated by sustaining an average flux at peak flow, through an increase in membrane surface area, or by temporarily increasing flux during periods of peak flow. This latter option will constrain capital investment in membrane surface area by up to three times, but its viability is impingent upon permeability not being compromised in the long-term from the short-term turn-up in flux.
A peak ratio of 1.4–1.5 is recommended for full-scale aerobic MBR which assumes that a maximum sustainable flux (defined as the flux required to limit fouling and avoid or limit the demand for reactive chemical cleaning) can be achieved during peak flow that is 40–50% higher than the average flux [7], [8], [9]. Some full-scale aerobic MBR plants have adopted more conservative design, instead specifying the membrane surface area to match peak flow, which ensures a considerably lower operating flux during flow variation [8], [10], but introduces a tremendous penalty in capital cost. This is significant since it is estimated that membrane area will comprise the largest proportion of capital cost (61–72%) for a full-scale municipal wastewater AnMBR [11], [12]. Furthermore, by specifying membrane surface area based on peak flow, severe membrane under-utilisation has been reported [8]. To illustrate, in several surveys of full-scale municipal aerobic MBRs [13], [14], the average flow was typically less than 50% of the peak flow used for design. This also incurred an increased operational cost of around 54%, due to the excess specific aeration demand per unit membrane area (SADm) required [8]. In the context of AnMBR for municipal wastewater treatment, this increase in energy demand and operational cost may reduce the attractiveness of investment, since the core aspiration is to facilitate energy neutral wastewater treatment [15].
Whilst the implications of peak flow on AnMBR design and operation are yet to be reported, laboratory and pilot scale evaluation of aerobic MBR have been conducted, in which the capacity for the membrane to withstand an increase in flux, in response to peak flow, has been determined using a constant SADm [5], [16], [17]. Lebegue et al. [17] identified no significant difference in transmembrane pressure (TMP) before and after a 2 h peak flow event in a lab-scale aerobic MBR treating synthetic municipal wastewater, which increased flux from 10 to 30 L m−2 h−1 for two hours on a daily basis. However, Metcalf [9] observed a significant membrane permeability decline in a pilot scale aerobic MBR treating settled municipal wastewater, when the flux returned to the average flux of 20 L m−2 h−1, from a peak flux of 25 L m−2 h−1 that was sustained for 24 h. The authors attributed the increased fouling to the operating flux exceeding the critical flux during peak flow. In recognition of such behaviour, several studies sought to identify fouling control strategies that could be deployed during peak flow, such as increasing SADm, shortening filtration cycle time, or increasing backwash flux [4], [14]. Following evaluation of a laboratory scale aerobic MBR treating synthetic settled municipal wastewater, Howell et al. [18] concluded that membrane fouling introduced by a temporary increase in flux could be controlled by an increase in SADm, with the residual foulant removed following flux restoration to a sub-critical level. Hirani et al. [4] tested five different pilot-scale submerged aerobic MBRs treating settled municipal wastewater, and demonstrated that a reduction in membrane permeability of 22–32% following the introduction of a peak flux ratio 1.6–3.2, was reversible, indicating that the reactive implementation of physical cleaning strategies during peak flow, were effective to cope with peak flow [4]. Importantly, such observations suggest that membrane surface area can be specified based on average flow rather than peak flow, which would help constrain membrane capital investment.
In AnMBR, the bulk sludge matrix is considerably more complex than in conventional aerobic MBR, leading to significantly higher membrane fouling [7], [19]. As such, the reported flux for AnMBR is ordinarily between 5 and 12 L m−2 h−1 [1], [20], which is considerably below the flux of 20–30 L m−2 h−1 typically specified for full-scale aerobic MBR [7]. The membrane area required for AnMBR will therefore be greater than for aerobic MBR, with the membrane cost inevitably increasing when membrane area is specified to sustain average flux at peak flow. The aim of this paper is therefore to evaluate the impact of a temporary increase in AnMBR flux, in response to peak flow, to ascertain whether AnMBR membrane surface area can be specified based on average flow rather than peak flow in order to diminish capital investment. The specific objectives were to: (i) evaluate the parameters governing permeability recovery (initial flux, peak flux to initial flux ratio, peak length); (ii) investigate the impacts of peak flow and strategies of increased gas sparging during the peak to enhance permeability recovery; and (iii) compare the conventional and alternative hydrodynamic conditions, to sustain permeability recovery whilst minimising energy demand.
Section snippets
Anaerobic MBR pilot plant
The AnMBR pilot plant was configured as a granular upflow anaerobic sludge blanket (G-UASB) reactor with a submerged hollow fibre membrane cited downstream (Fig. 1). The UASB was 42.5 L in volume, and was fitted with a lamella plate clarifier for solid/liquid/gas separation (Paques, Balk, The Netherlands). Granular sludge (16 L) from a mesophilic UASB designed for the pulp and paper industry, was used for inoculum, and was left to acclimate for 360 days before experimentation commenced. Settled
Characterisation of AnMBR mixed liquor and critical flux determination
The mixed liquor within the membrane tank comprised MLSS and soluble microbial product (SMP) concentrations of 123 ± 38 mgSS L−1 and 90 ± 19 mgCOD L−1 respectively (Table 1). The critical flux (Jc) of the suspension was identified at two specific gas flow rates (Fig. 4). At a SGDm of 0.5 m3 m−2 h−1, the Jc was between 9 and 12 L m−2 h−1 and increased to between 12 and 15 L m−2 h−1, when SGDm was increased to 2.0 m3 m−2 h−1.
Impact of gas sparging on AnMBR membrane permeability following peak flow
The impact of peak flow (Qpeak, 2Q) on membrane permeability was
Discussion
In this study, the potential to restore permeability following peak flow has been evidenced in AnMBR treating municipal wastewater. Although future complementary research focussed on longer-term impacts of peak flow to permeability would be beneficial, data from this study suggests that the membrane area requirement for AnMBR can be potentially specified based on average flow instead of peak flow, manifesting in a considerable reduction in capital cost by about 67% compared with the design
Conclusions
The impact of critical transient peak flow characteristics (peak duration, frequency and size) on membrane permeability has been evaluated, together with several reactive methods to improve permeability recovery following peakflow events. The following conclusions can be drawn:
- •
Enhanced permeability recovery is achieved by increasing gas sparging during peak flow. However, considerable increase in gas sparging is needed to shift the critical flux of the suspension (four times), leading to a
Acknowledgements
The authors would like to thank our industrial sponsors Anglian Water, Scottish Water, Severn Trent Water and Thames Water for their financial and technical support.
References (41)
- et al.
Impact on reactor configuration on the performance of anaerobic MBRs: treatment of settled sewage in temperate climates
Water Res.
(2013) - et al.
Pilot plant study of a new rotating hollow fibre membrane module for improved performance of an anaerobic submerged MBR
J. Membr. Sci.
(2016) - et al.
Factors that affect the permeability of commercial hollow-fibre membranes in a submerged anaerobic MBR (HF-SAnMBR) system
Water Res.
(2013) - et al.
Peak flux performance and microbial removal by selected membrane bioreactor systems
Water Res.
(2010) - et al.
The cost of a large-scale hollow fibre MBR
Water Res.
(2010) - et al.
Feasibility evaluation of submerged anaerobic membrane bioreactor for municipal secondary wastewater treatment
Desalination
(2011) - et al.
MBR functioning under steady and unsteady state conditions. Impact on performances and membrane fouling dynamics
Desalination
(2008) - et al.
In situ manipulation of critical flux in a submerged membrane bioreactor using variable aeration rates, and effects of membrane history
J. Membr. Sci.
(2004) - et al.
Impact of membrane configuration on fouling in anaerobic membrane bioreactors
J. Membr. Sci.
(2011) - et al.
The performance enhancements of upflow anaerobic sludge blanket (UASB) reactors for domestic sludge treatment – a state-of-the-art review
Water Res.
(2012)
Comparison of dead-end and continuous filtration conditions in a denitrification membrane bioreactor
J. Membr. Sci.
Critical flux determination by the flux-step method in a submerged membrane bioreactor
J. Membr. Sci.
Protein measurement with the folin phenol reagent
J. Biol. Chem.
Optimisation of dead-end filtration conditions for an immersed anoxic membrane bioreactor
J. Membr. Sci.
Flux criticality and sustainability in a hollow fibre submerged membrane bioreactor for municipal wastewater treatment
J. Membr. Sci.
Sub-critical filtration conditions of commercial hollow-fibre membranes in a submerged anaerobic MBR (HF-SAnMBR) system: the effect of gas sparging intensity
Bioresour. Technol.
Investigations of cake fouling during the cross-flow microfiltration of a model suspension: influence of buoyancy on deposition and shear-induced removal
Procedia Eng.
Analysis of fouling mechanisms in anaerobic membrane bioreactors
Water Res.
The influence of dwelling occupancy and day of the week on domestic appliance wastewater discharges
Build. Environ.
A new approach to backwash initiation in membrane systems
J. Membr. Sci.
Cited by (21)
The third route: A techno-economic evaluation of extreme water and wastewater decentralization
2022, Water ResearchCitation Excerpt :anMBRs modeled in this study were sized according to their HRT. The low-strength wastewater (i.e., greaywater) could be successfully treated using anMBR technology (Gouveia et al., 2015; Martin Garcia et al., 2013; Shin et al., 2014; Wang et al., 2018b, 2018a). It must to be noted, that as anMBRs achieve a high solids separation, the clogging of the pores is not expected to be major concern (Membrana, Wuppertal, Germany); microporous contactors designed for low flow rates (instead than non-porous membranes) appear to be the best approaches for higher methane recovery yields.
Anaerobic membrane bioreactors (AnMBRs) for municipal wastewater treatment- potential benefits, constraints, and future perspectives: An updated review
2022, Science of the Total EnvironmentCitation Excerpt :The energy demand of AnMBR was observed to be strongly influenced by membrane operating flux and SGD. Wang et al. (2018b) measured the performance of AnMBR during the unsteady-steady state for the treatment of municipal wastewater and observed that increased gas sparging was effective for peak flow which required 0.04 kWh/m3 of energy for pseudo-dead-end filtration for gas sparging. This particular membrane fouling control also enabled higher fluxes in other MBR applications as well even in peak flow scenarios over longer periods of time.
Anaerobic membrane bioreactors (AnMBR) treating urban wastewater in mild climates
2020, Bioresource TechnologyCitation Excerpt :Although AnMBR technology is currently used to treat industrial effluents (Dereli et al., 2012), full scale applications for UWW have not yet been reported (Shin and Bae, 2018). Various authors have evaluated the performance of different AnMBR pilots (e.g. Aslam et al., 2017; Giménez et al., 2011; Gouveia et al., 2015; Kim et al., 2011; Li and Wang, 2006; Martin Garcia et al., 2013; Martinez-Sosa et al., 2011; Shin et al., 2014; Wang et al., 2018). However, it is essential to increase Technology Readiness Level (TRL) to demonstrate the feasibility of AnMBR technology for full-scale UWW treatment.
Advances in anaerobic membrane bioreactor technology for municipal wastewater treatment: A 2020 updated review
2020, Renewable and Sustainable Energy ReviewsCitation Excerpt :Overall, the development of FO membranes able to achieve high water fluxes and low reverse solute fluxes is crucial to reduce capital and operating costs of AnOMBRs. Most common AnMBR-UASB configurations include the membrane module in an external tank [49,98,117,144,145] or at the top of the UASB reactor [77,146,147]. The former is the most reported system.