Scale-up of reverse electrodialysis for energy generation from high concentration salinity gradients.

Whilst reverse electrodialysis (RED) has been extensively characterised for saline gradient energy from seawater/river water (0.5 M/0.02 M), less is known about RED stack design for high concentration salinity gradients (4 M/0.02 M), important to closed loop applications (e.g. thermal-to-electrical, energy storage). This study therefore focuses on the scale-up of RED stacks for high concentration salinity gradients. Higher velocities were required to attain a maximum Open Circuit Voltage (OCV) for 4 M/0.02 M, which gives a measure of the electrochemical potential of the cell. The experimental OCV was also much below the theoretical OCV, due to the greater boundary layer resistance observed, which is distinct from 0.5 M/0.02 M. However, negative net power density (net produced electrical power divided by total membrane area) was demonstrated with 0.5 M/0.02 M for larger stacks using shorter residence times (three stack sizes tested: 10 × 10cm, 10 × 20cm and 10 × 40cm). In contrast, the highest net power density was observed at the shortest residence time for the 4 M/0.02 M concentration gradient, as the increased ionic flux compensated for the pressure drop. Whilst comparable net power densities were determined for the 10 × 10cm and 10 × 40cm stacks using the 4 M/0.02 M concentration gradient, the osmotic and ionic transport mechanisms are distinct. Increasing cell pair number improved maximum current density. This subsequently increased power density, due to the reduction in boundary layer resistance, and may therefore be used to improve thermodynamic efficiency and power density from RED for high concentrations. Although comparable power densities may be achieved for small and large stacks, large stacks maybe preferred for high concentration salinity gradients due to the comparative benefit in thermodynamic efficiency in single pass. The greater current achieved by large stacks may also be complemented by an increase in cell pair number and current density optimisation to increase power density and reduce exergy losses.

Transitioning from electrodialysis to reverse electrodialysis stack design for energy generation from high concentration salinity gradients.

In this study, stack design for high concentration gradient reverse electrodialysis operating in recycle is addressed. High concentration gradients introduce complex transport phenomena, which are exacerbated when recycling feeds; a strategy employed to improve system level energy efficiency. This unique challenge indicates that membrane properties and spacer thickness requirements may differ considerably from reverse electrodialysis for lower concentration gradients (e.g. seawater/river water), drawing closer parallels to electrodialysis stack design. Consequently, commercially available electrodialysis and reverse electrodialysis stack design was first compared for power generation from high concentration gradients. Higher gross power densities were identified for the reverse electrodialysis stack, due to the use of thinner membranes characterised by a higher permselectivity, which improved current. However, energy efficiency of the electrodialysis stack was twice that recorded for the reverse electrodialysis stack at low current densities, which was attributed to: (i) an increased residence time provided by the larger intermembrane distance, and (ii) reduced exergy losses of the electrodialysis membranes, which provided comparatively lower water permeance. Further in-depth investigation into membrane properties and spacer thickness identified that membranes characterised by an intermediate water permeability and ohmic resistance provided the highest power density and energy efficiency (Neosepta ACS/CMS), while wider intermembrane distances up to 0.3 mm improved energy efficiency. This study confirms that reverse electrodialysis stacks for high concentration gradients in recycle therefore demand design more comparable to electrodialysis stacks to drive energy efficiency, but when selecting membrane properties, the trade-off with permselectivity must also be considered to ensure economic viability.

Managing power dissipation in closed-loop reverse electrodialysis to maximise energy recovery during thermal-to-electric conversion.

Whilst the efficiency of reverse electrodialysis (RED) for thermal-to-electrical conversion has been theoretically demonstrated for low-grade waste heat, the specific configuration and salinity required to manage power generation has been less well described. This study demonstrates that operating RED by recycling feed solutions provides the most suitable configuration for energy recovery from a fixed solution volume, providing a minimum unitary cost for energy production. For a fixed membrane area, recycling feeds achieves energy efficiency seven times higher than single pass (conventional operation), and with an improved power density. However, ionic transport, water flux and concentration polarisation introduce complex temporal effects when concentrated brines are recirculated, that are not ordinarily encountered in single pass systems. Regeneration of the concentration gradient at around 80% energy dissipation was deemed most economically pragmatic, due to the increased resistance to mass transport beyond this threshold. However, this leads to significant exergy destruction that could be improved by interventions to better control ionic build up in the dilute feed. Further improvements to energy efficiency were fostered through optimising current density for each brine concentration independently. Whilst energy efficiency was greatest at lower brine concentrations, the work produced from a fixed volume of feed solution was greatest at higher saline concentrations. Since the thermal-to-electrical conversion proposed is governed by volumetric heat utilisation (distillation to reset the concentration gradient), higher brine concentrations are therefore recommended to improve total system efficiency. Importantly, this study provides new evidence for the configuration and boundary conditions required to realise RED as a practical solution for application to sources of low-grade waste heat in industry.

Hybrid membrane distillation reverse electrodialysis configuration for water and energy recovery from human urine: An opportunity for off-grid decentralised sanitation.

The integration of membrane distillation with reverse electrodialysis has been investigated as a sustainable sanitation solution to provide clean water and electrical power from urine and waste heat. Reverse electrodialysis was integrated to provide the partial remixing of the concentrate (urine) and diluate (permeate) produced from the membrane distillation of urine. Broadly comparable power densities to those of a model salt solution (sodium chloride) were determined during evaluation of the individual and combined contribution of the various monovalent and multivalent inorganic and organic salt constituents in urine. Power densities were improved through raising feed-side temperature and increasing concentration in the concentrate, without observation of limiting behaviour imposed by non-ideal salt and water transport. A further unique contribution of this application is the limited volume of salt concentrate available, which demanded brine recycling to maximise energy recovery analogous to a battery, operating in a 'state of charge'. During recycle, around 47% of the Gibbs free energy was recoverable with up to 80% of the energy extractable before the concentration difference between the two solutions was halfway towards equilibrium which implies that energy recovery can be optimised with limited effect on permeate quality. This study has provided the first successful demonstration of an integrated MD-RED system for energy recovery from a limited resource, and evidences that the recovered power is sufficient to operate a range of low current fluid pumping technologies that could help deliver off-grid sanitation and clean water recovery at single household scale.

Modelling the energy demands of aerobic and anaerobic membrane bioreactors for wastewater treatment.

A modelling study has been developed in which the energy requirements of aerobic and anaerobic membrane bioreactors (MBRs) are assessed in order to compare these two wastewater treatment technologies. The model took into consideration the aeration required for biological oxidation in aerobic MBRs (AeMBRs), the energy recovery from methane production in anaerobic MBRs (AnMBRs) and the energy demands of operating submerged and sidestream membrane configurations. Aeration and membrane energy demands were estimated based on previously developed modelling studies populated with operational data from the literature. Given the difference in sludge production between aerobic and anaerobic systems, the model was benchmarked by assuming high sludge retention times or complete retention of solids in both AeMBRs and AnMBRs. Analysis of biogas production in AnMBRs revealed that the heat required to achieve mesophilic temperatures (35 degrees C) in the reactor was only possible with influent wastewater strengths above 4-5 g COD L(-1). The general trend of the submerged configuration, which is less energy intensive than the sidestream configuration in aerobic systems, was not observed in AnMBRs, mainly due to the wide variation in gas demand utilized in anaerobic systems. Compared to AeMBRs, for which the energy requirements were estimated to approach 2 kWh m(-3) (influent up to 1 g COD L(-1)), the energy demands associated with fouling control in AnMBRs were lower (0.80 kWh m(-3) for influent of 1.14 g COD L(-1)), although due to the low fluxes reported in the literature capital costs associated with membrane material would be three times higher than this.

Comparison of grey water treatment performance by a cascading sand filter and a constructed wetland.

A novel unplanted vertical flow subsurface constructed wetland technology comprising three shallow beds (0.6 m length, 0.45 m width and 0.2 m depth) arranged in a cascading series and a standard single-pass Vertical Flow Planted Constructed Wetland (VFPCW, 6 m² and 0.7 m depth) were tested for grey water treatment. Particular focus was on meeting consent for published wastewater reuse parameters and removal of anionic surfactants. Treatment performance at two hydraulic loading rates (HLR) of 0.08, and 0.17 m³ m⁻² d⁻¹ were compared. Both technologies effectively removed more than 90% turbidity and more than 96% for organics with the prototype meeting the most stringent reuse standard of < 2 NTU and <10 mg/L. However, surfactant removal in the VFPCW was higher (76-85%) than in the prototype which only achieved more than 50% removal at higher loading rate. Generally, the prototype performed consistently better than the VFPCW except for surfactant removal. However, at higher loading rates, both systems did not meet the reuse standard of <1 mg L⁻¹ for anionic surfactants. This observation confirms that shallow beds provide a more oxidised environment leading to higher BOD5 and COD removals. Presence of plants in the VFPCW led to higher anionic surfactant removal, through increased microbial and sorption processes.

Membrane chemical reactor (MCR) combining photocatalysis and microfiltration for grey water treatment.

Urban water recycling is now becoming an important issue where water resources are becoming scarce. This paper looks at reusing grey water; the preference is treatment processes based on biological systems to remove the dissolved organic content. Here, an alternative process, photocatalysis is discussed as it is an attractive technology that could be well-suited for treating the recalcitrant organic compounds found in grey water. The photocatalytic process oxidises organic reactants at a catalyst surface in the presence of ultraviolet light. Given enough exposure time, organic compounds will be oxidized into CO2 and water. The best contact is achieved in a slurry reactor but a second step to separate and recover the catalyst is need. This paper discusses a new membrane chemical reactor (MCR) combining photocatalysis and microfiltration for grey water treatment.

Impact on reactor configuration on the performance of anaerobic MBRs: treatment of settled sewage in temperate climates.

The treatment efficiency and membrane performance of a granular and suspended growth anaerobic membrane bioreactor (G-AnMBR and AnMBR respectively) were compared and evaluated. Both anaerobic MBRs were operated in parallel during 250 days with low strength wastewater and under UK weather conditions. Both systems presented COD and BOD removal efficiencies of 80-95% and >90% respectively. Effluent BOD remained between 5 and 15 mgBOD L(-1) through the experimental period while effluent COD increased from 25 mg L(-1) to 75 mg L(-1) as temperature decreased from 25 °C to 10 °C respectively indicating the production of non biodegradable organics at lower temperatures. Although similar levels of low molecular weight organics were present in the sludge supernatant, recycling of the mixed liquor from the membrane tank to the bioreactor at a low upflow velocity enhanced interception of solids in the sludge bed of the G-AnMBR limiting the solid and colloidal load to the membrane as compared to the suspended system. Results from flux step test showed that critical flux increased from 4 to 13 L m(-2) h(-1) and from 3 to 5 L m(-2) h(-1) with gas sparging intensities varying from 0.007 m s(-1) to 0.041. Additional long term trials in which the effect of gas sparging rate and backwashing efficiency were assessed confirmed the lower fouling propensity of the G-AnMBR.

Microbial community analysis of fouled reverse osmosis membranes used in water recycling.

Biofouling on RO membranes has major cost implications in water reclamation. In this study membranes and water samples were collected from a RO pilot-plant operated on two sites to study the differences in microbial communities in order to develop a better understanding of the biofouling. For the two sites studied, the examination of the front membrane of the first stage and the tail membrane of the second stage of the RO train using 16S rRNA gene-based molecular technique showed that bacteria were similar on both stages and no significant effect of the membrane location within the RO train on the biofilm development could be discerned. However, the comparison of the identified bacteria from membrane samples between the two sites showed that each site is specific, leading to a different composition of microbial communities. The different nutrient concentrations in the RO feed water due to the different biological pre-treatments are one potential explanation for the observed differences in the microbial communities. Seasonal variations also play a major role in the development of microbial communities as shown by the significant differences observed between the communities measured in the samples in winter and summer on the second site. The results did not show similarity between the species identified on the RO membranes and in the feed water. Hence, the relationship of microbial community between the water generated during the pre-treatment process and RO membranes is not obvious. From this study, results showed that there is an actual need to investigate the development of microbial communities on membrane surface in real conditions in order to suggest tailored solutions for biofouling control and removal.

Impact of effluent organic matter on low-pressure membrane fouling in tertiary treatment.

This study aims at comparing low-pressure membrane fouling obtained with two different secondary effluents at bench and pilot-scale based on the determination of two fouling indices: the total fouling index (TFI) and the hydraulically irreversible fouling index (HIFI). The main objective was to investigate if simpler and less costly bench-scale experimentation can substitute for pilot-scale trials when assessing the fouling potential of secondary effluent in large scale membrane filtration plants producing recycled water. Absolute values for specific flux and total fouling index for the bench-scale system were higher than those determined from pilot-scale, nevertheless a statistically significant correlation (r(2) = 0.63, α = 0.1) was obtained for the total fouling index at both scales. On the contrary no such correlation was found for the hydraulically irreversible fouling index. Advanced water characterization tools such as excitation-emission matrix fluorescence spectroscopy (EEM) and liquid chromatography with organic carbon detection (LC-OCD) were used for the characterization of foulants. On the basis of statistical analysis, biopolymers and humic substances were found to be the major contribution to total fouling (r(2) = 0.95 and r(2) = 0.88, respectively). Adsorption of the low molecular weight neutral compounds to the membrane was attributed to hydraulically irreversible fouling (r(2) = 0.67).