Resilience towards organic load and activated sludge variations in co-fermentation for carboxylic acid production.

Two perturbations were investigated in acidogenic co-fermentation of waste activated sludge (WAS) and food waste in continuous mesophilic fermenters: increasing the organic loading rate (OLR) and changing the WAS. A control reactor maintained an OLR of 11 gVS/(L·d), while a test reactor had a prolonged OLR change to 18 gVS/(L·d). For each OLR, two WAS were studied. The change in OLR led to differentiated fermentation product profile without compromising the fermentation yields (∼300 mgCOD/gVS). At 11 gVS/(L·d), the product profile was dominated by acetic, butyric, and propionic acids while at 18 gVS/(L·d) it shifted to acetic acid, ethanol, and caproic acid. Reverting the OLR also reverted the fermentation profile. The biomass immigration with the WAS changed the fermentation microbial structure and introduced acetic acid-consuming methanogens, which growth was only delayed by the OLR increase. Microbial monitoring and post-fermentation tests can be used for early detection of acetic acid-consuming events.

Fractional calculus as a generalized kinetic model for biochemical methane potential tests.

This study presents a fractional calculus model as a generalized kinetic model for estimating the maximum methane yield and degradation kinetics in biomethane potential (BMP) assays, a key analytical method in anaerobic digestion research and application. The fractional model outperformed common first-order kinetic models by yielding superior data fitting and properly managing substrate heterogeneity. The fractional model showed robust performance in mono-digestion, co-digestion and pre-treatment BMP assays with or without presence of large tailing or sigmoidal patterns in the BMP curve. The main advantage of the fractional model over other models is its ability to capture the complexities of the methane production process without losing model accuracy. Assessment of the mathematical model revealed that for fractional orders greater than 0.8 the Mittag-Leffler sequence could be transformed into a more computationally efficient exponential function.

Temperature-driven carboxylic acid production from waste activated sludge and food waste: Co-fermentation performance and microbial dynamics.

This work aims to improve the continuous co-fermentation of waste activated sludge (WAS) and food waste (FW) by investigating the long-term impact of temperature on fermentation performance and the underpinning microbial community. Acidogenic co-fermentation of WAS and FW (70:30 % VS-basis) to produce volatile fatty acids (VFA) was studied in continuous fermenters at different temperatures (25, 35, 45, 55 °C) at an organic loading rate of 11 gVS/(L·d) and a hydraulic retention time of 3.5 days. Two batches of WAS (A and B) were collected from the same wastewater treatment plant at different periods to understand the impact of the WAS microbioota on the fermenters' microbial communities. Solubilisation yield was higher at 45 °C (575 ± 68 mgCOD/gVS) followed by 55 °C (508 ± 45 mgCOD/gVS). Fermentation yield was higher at 55 °C (425 ± 28 mgCOD/gVS) followed by 35 °C (327 ± 17 mgCOD/gVS). Temperature also had a noticeable impact on the VFA profile. At 55 °C, acetic (40 %) and butyric (40 %) acid dominated, while acetic (37 %), butyric acid (31 %), and propionic acid (17 %) dominated at 35 °C. At 45 °C, an accumulation of caproic acid was detected which did not occur at other temperatures. Each temperature had a distinct microbial community, where the WAS microbiota played an important role. The biomass mass-balance showed the highest growth of microorganisms (51 %) at 35 °C and WAS_B, where a consumption of acetic acid was observed. Therefore, at 35 °C, there is a higher risk of acetic acid consumption probably due to the proliferation of methanogens imported from WAS.

Impact of food waste composition on acidogenic co-fermentation with waste activated sludge.

The impact of food waste (FW) composition on co-fermentation performance was studied to elucidate if adjusting FW composition can be used to drive the fermentation yield and profile, which is relevant for biorefinery applications. First, the impact of individual FW components (i.e., fruit, vegetables, pasta, rice, meat, fish, and cellulose) was assessed. Subsequently, the effect of mixing a protein-rich component and a carbohydrate-rich component was studied (i.e., fish/fruit and fish/cellulose, and meat/rice and meat/vegetable). All experiments were carried out in mesophilic batch assays using waste activated sludge (WAS) as main substrate, the same mixture ratio (70 % WAS +30 % FW on VS basis), and no pH control. Results showed that each FW component had a distinct effect on VFA yield and profile, with protein-rich components reaching the highest VFA yields; 502 and 442 mgCOD/gVS for WAS/Fish and WAS/Meat, respectively. A positive interaction on VFA yield was observed when mixing a protein-rich and a carbohydrate-rich component. This interaction was not proportional to the co-substrates proportion in the mixtures. On the other hand, the VFA profile was clearly driven by the components in the mixture, including both WAS and FW composition. Overall, these results indicate that predicting the VFA yield of WAS/FW co-fermentation is not just related to FW composition, but FW composition could be used to adjust the VFA profile to a certain extent.

Gamma distribution function to understand anaerobic digestion kinetics: Kinetic constants are not constant.

The Gamma model is a novel approach to characterise the complex degradation dynamics taking place during anaerobic digestion. This three parameters model results from combining the first-order kinetic model and the Gamma distribution function. In contrast to conventional models, where the kinetic constant is considered invariant, the Gamma model allows analysing the variability of the kinetic constant using a probability density function. The kinetic constant of mono-digestion and co-digestion batch tests of different wastes were modelled using the Gamma model and two common first-order models: one-step one-fraction model and one-step two-fraction model. The Gamma distribution function approximates three distinct probability density functions, i.e. exponential, log-normal, and delta Dirac. Specifically, (i) cattle paunch and pig manure approximated a log-normal distribution; (ii) cattle manure and microalgae approximated an exponential distribution, and (iii) primary sludge and cellulose approximated a delta Dirac distribution. The Gamma model was able to characterise two distinct waste activated sludge, one approximated to a log-normal distribution and the other to an exponential distribution. The same cellulose was tested with two different inocula; in both tests, the Gamma distribution function approximated a delta Dirac function but with a different kinetic value. The potential and consistency of Gamma model were also evident when analysing pig manure and microalgae co-digestion batch tests since (i) the mean k of the co-digestion tests were within the values of the mono-digestion tests, and (ii) the profile of the density function transitioned from log-normal to exponential distribution as the percentage of microalgae in the mixture increased.

Ammonia recovery from acidogenic fermentation effluents using a gas-permeable membrane contactor.

A gas-permeable membrane (GPM) contactor was used to recover ammoniacal nitrogen from a synthetic and a biowaste fermentation broth under different pH (from 6 to 11) and temperatures (35 and 55 °C). Ammonia mass transfer constant (Km) increased as pH and temperature increased. For synthetic broth, pH 10 provided the best results, when considering the Km (9.2·10-7 m·s-1) and the reagents consumption (1.0 mol NaOH·mol-1 TAN and 0.6 mol H2SO4·mol-1 TAN). Biowaste fermentation generated a broth with a high concentration of ammoniacal nitrogen (4.9 g N·L-1) and volatile fatty acids (VFA) (41.1 g COD·L-1). Experiments using the biowaste broth showed a lower Km (5.0·10-7 m·s-1 at pH 10) than the synthetic broth, related to the solution matrix and other species interference. VFAs were not detected in the trapping solution. Overall, these results show that GPM is a suitable technology to efficiently separate ammoniacal nitrogen and VFA from fermentation broths.

Potential of anaerobic co-fermentation in wastewater treatments plants: A review.

Fermentation (not anaerobic digestion) is an emerging biotechnology to transform waste into easily assimilable organic compounds such as volatile fatty acids, lactic acid and alcohols. Co-fermentation, the simultaneous fermentation of two or more waste, is an opportunity for wastewater treatment plants (WWTPs) to increase the yields of sludge mono-fermentation. Most publications have studied waste activated sludge co-fermentation with food waste or agri-industrial waste. Mixing ratio, pH and temperature are the most studied variables. The highest fermentation yields have been generally achieved in mixtures dominated by the most biodegradable substrate at circumneutral pH and mesophilic conditions. Nonetheless, most experiments have been performed in batch assays which results are driven by the capabilities of the starting microbial community and do not allow evaluating the microbial acclimation that occurs under continuous conditions. Temperature, pH, hydraulic retention time and organic load are variables that can be controlled to optimise the performance of continuous co-fermenters (i.e., favour waste hydrolysis and fermentation and limit the proliferation of methanogens). This review also discusses the integration of co-fermentation with other biotechnologies in WWTPs. Overall, this review presents a comprehensive and critical review of the achievements on co-fermentation research and lays the foundation for future research.

Nitrogen recovery from pig slurry by struvite precipitation using a low-cost magnesium oxide.

Ammonia nitrogen management is a recurrent problem in intensive livestock areas. Struvite precipitation stands as a mature technology to recover ammonia nitrogen and prevent associated environmental problems. However, the feasibility of struvite technology to recover ammonia nitrogen from pig manure is limited by the reagents cost. This research aimed to optimise the formulation of a stabilizing agent (SA) synthesised using an industrial low-grade MgO by-product (LG-MgO) and phosphoric acid for efficient TAN recovery via struvite precipitation. Experimental results showed that the H3PO4/LG-MgO ratio controls the magnesium phosphate mineral phase of the SA (bobierrite and/or newberyite). Newberyite-rich SA showed the highest TAN removal efficiency from pig manure (66-73%) compared to the SA formed by a mixture of newberyite and bobierrite (51-59%) and by bobierrite (26%). Particle size reduction of LG-MgO did not improve the SA's TAN removal efficiency, although XRD patterns showed that the precipitates from the TAN removal experiments contained some unreacted newberyite. The economic analysis showed that the higher reactivity of the SA formulated using higher H3PO4/LG-MgO ratios compensated reagent costs. The SA synthesised with a H3PO4/LG-MgO ratio of 0.98 showed the most economical treatment cost, which was estimated at 7.5 € per kg of ammonia nitrogen from pig manure. Finally, the optimum SA was successfully synthesised in a 200-L pilot plant, with a TAN removal capacity only 10% lower than the one synthesised at lab-scale.

Assessing the potential of waste activated sludge and food waste co-fermentation for carboxylic acids production.

This study investigated waste activated sludge (WAS) and food waste (FW) co-fermentation in batch assays to produce carboxylic acids. Three mixtures (50%, 70% and 90% WAS in VS basis) were studied under different conditions: with and without extra alkalinity, and with and without WAS auto-hydrolysis pre-treatment. All tests were carried out at 35 °C, without pH adjustment and without external inoculum. Experimental results showed that co-fermentation yields, including volatile fatty acids and lactic acid, were always higher than WAS and FW mono-fermentation yields (ca. 100 and 80 mgCOD/gVS, respectively). Co-fermentation yields increased as the proportion of FW in the mixture increased, indicating that the improvement was primarily due to a higher FW degradation under co-fermentation conditions. The maximum co-fermentation yield was on average 480 mgCOD/gVS for the WAS/FW_50/50 mixture. The importance of pH on co-fermentation performance was evident in the experiments carried out with extra alkalinity, which showed that the proportion of WAS in the mixture should be high enough to keep the pH above 5.0. However, fermenters operational conditions should also prevent the enrichment of acetic acid consuming microorganisms. WAS auto-hydrolysis pre-treatment did not enhance co-fermentation yields but showed minor kinetic improvements. Regarding the product profile, butyric acid was enriched as the proportion of FW in the mixture increased and the concomitant pH decreased to the detriment of propionic acid. Propionic acid prevailed under neutral pH in the WAS mono-fermentation and the WAS/FW_90/10 mixture.

Simultaneous ammonium and phosphate recovery and stabilization from urban sewage sludge anaerobic digestates using reactive sorbents.

The use of low-cost inorganic sorbents as a new sustainable strategy to enhance the valorization of nutrients (N-P-K), from the urban water cycle (e.g., side streams from sewage sludge anaerobic digestion), in agriculture applications is presented. The simultaneous recovery and stabilization of ammonium and phosphate by using a mixture of two reactive sorbents (Na and K zeolites and magnesium oxide) was evaluated. The nutrients stabilization process, favoured at alkaline pH values, is carried out by a) the precipitation of phosphate ions with magnesium and/or ammonium ions and b) the sorption of ammonium by Na- and K-zeolites. MgO(s) promoted the stabilization of phosphate as bobierrite (Mg3(PO4)2(s)) or struvite (MgNH4PO4(s)) depending on the applied dose. Doses with the stoichiometric molar ratio of Mg/P promote the formation of bobierrite, while molar ratios higher than 3 favour the formation of struvite. Na zeolites (NaP1-NA, NaP1-IQE) demonstrated efficiency on ammonium stabilization between 60±2 (for 15gZ/L) to 90±3% (for 50gZ/L). The ammonium recovery efficiency is limited by the zeolite sorption capacity. If the target of the fertilizing criteria should include K, then the use of a K-zeolite (e.g., 5AH-IQE) provides a good solution. The optimum pH for the precipitation of struvite and bobierrite is 9.5 and the optimum pH for ammonium removal is between 4 and 8.5. N is present in higher concentrations (up 0.7-1gNH4+/L) when pH is ranged between 8.2 and 8.6. The ammonium recovery ratios were better than those previously reported using only magnesium oxide or even a more expensive reagent as newberrite (MgHPO4(s)). The recovery mechanisms described generate low-solubility stabilized nutrients forms that potentially can be applied as slow-release fertilizers in agriculture. Thus, the use in agriculture of blends of digested sludge with low-solubility stabilized nutrients forms will improve soils quality properties in terms of organic matter and nutrients availability.

Impact of paper and cardboard suppression on OFMSW anaerobic digestion.

Mechanical-biological treatment plants treat municipal solid waste to recover recyclable materials, nutrients and energy. Waste paper and cardboard (WP), the second main compound in municipal solid waste (∼30% in weight basis), is typically used for biogas generation. However, its recovery is gaining attention as it can be used to produce add-value products like bioethanol and residual derived fuel. Nevertheless, WP suppression or replacement will impact anaerobic digestion in terms of biogas production, process stability and digestate management. Two lab-scale reactors were used to assess the impact of WP in anaerobic digestion performance. A control reactor was only fed with biowaste (BioW), while a second reactor was fed with two different mixtures of BioW and WP, i.e. 85/15% and 70/30% (weight basis). Results indicate that either replacing half of the WP by BioW or removing half of the WP has little impact on the methane production. When removing half of the WP, methane production could be sustained by a larger waste biodegradability. The replacement of all WP by BioW increased the reactor methane production (∼37%), while removing all WP would have reduced the methane production about 15%. Finally, replacing WP loading rate by BioW led to a system less tolerant to instability periods and with poorer digestate quality.

Integrating the selection of PHA storing biomass and nitrogen removal via nitrite in the main wastewater treatment line.

A novel scheme was developed for the treatment of municipal wastewater integrating nitritation/denitritation with the selection of polyhydroxyalkanoates (PHA) storing biomass under an aerobic/anoxic, feast/famine regime. The process took place in a sequencing batch reactor (SBR) and the subsequent PHA accumulation in a batch reactor. The carbon source added during the selection and accumulation steps consisted of fermentation liquid from the organic fraction of municipal solids waste (OFMSW FL) (Period I) and OFMSW and primary sludge fermentation liquid (Period II). Selection of PHA storing biomass was successful and denitritation was driven by internally stored PHA during the famine phase. Under optimum conditions of SBR operation ammonia removal was 93%, reaching a maximum nitrite removal of 98%. The treated effluent met the nitrogen limits, while PHA-storing biomass was successfully selected. The maximum accumulation of PHA was 10.6% (wt.) since the nutrients present in the carbon source promoted bacterial growth.

Two-step partial nitritation/Anammox process in granulation reactors: Start-up operation and microbial characterization.

A two-stage Partial Nitritation (PN)/Anammox process was carried out at lab-scale conditions to treat reject water from a municipal WWTP. PN was achieved in a granular SBR obtaining an effluent with a NH4(+)-N/NO2(-)-N molar ratio around 1.0. The microbial characterization of this reactor revealed a predominance of Betaproteobacteria, with a member of Nitrosomonas as the main autotrophic ammonium oxidizing bacterium (AOB). Nitrite oxidizing bacteria (NOB) were under the detection limit of 16S rRNA gene pyrosequencing, indicating their effective inhibition. The effluent of the PN reactor was fed to an Anammox SBR where stable operation was achieved with a NH4(+)-N:NO2(-)-N:NO3(-)-N stoichiometry of 1:1.25:0.14. The deviation to the theoretical stoichiometry could be attributed to the presence of heterotrophic biomass in the Anammox reactor (mainly members of Chlorobi and Chloroflexi). Planctomycetes accounted for 7% of the global community, being members of Brocadia (1.4% of the total abundance) the main anaerobic ammonium oxidizer detected.

Aerobic/anoxic post-treatment of anaerobically digested sewage sludge as an alternative to biological nitrogen removal from reject water.

Stabilisation and biological nitrogen removal (BNR) of anaerobically digested sewage sludge were studied in a post-aeration reactor at pilot scale working under alternating anoxic-aerobic conditions. Digested sludge came from a two-stage anaerobic digestion (thermophilic + mesophilic). The best post-aerator performance was achieved when working at an HRT of 10 days (4 days aerobic; dissolved oxygen of 1.8 mg L(-1)) and VS content in the feed no lower than 6.7 g L(-1). Free ammonia concentration values in the effluent above 1.5 mg N L(-1) (around 150 mg NH4 (+)-N L(-1) at pH 7) were necessary to promote the BNR over nitrite. Removal efficiencies up to 80 % NH4 (+)-N, 50-55 % total nitrogen and 15-20 % VS were recorded in this study, with no external addition of chemicals. A nitrogen mass balance revealed that the high percent of NH4 (+)-N assimilated in heterotrophic growth was counteracted with that released in ammonification and fermentation, leading to a NH4 (+)-N removal mainly related to biological nitritation/denitritation.

Comparison of aerobic granulation and anaerobic membrane bioreactor technologies for winery wastewater treatment.

An anaerobic membrane bioreactor and aerobic granulation technologies were tested at laboratory scale to treat winery wastewater, which is characterised by a high and variable biodegradable organic load. Both technologies have already been tested for alcohol fermentation wastewaters, but there is a lack of data relating to their application to winery wastewater treatment. The anaerobic membrane bioreactor, with an external microfiltration module, was started up for 230 days, achieving a biogas production of up to 0.35 L CH4L(-1)d(-1) when 1.5 kg COD m(-3)d(-1) was applied. Average flux was 10.5 L m(-2) h(-1) (LMH), obtaining a treated effluent free of suspended solids and a chemical oxygen demand (COD) concentration lower than 100 mg COD L(-1). In contrast, the aerobic granular sequencing batch reactor coped with 15 kg COD m(-3)d(-1), but effluent quality was slightly worse. Aerobic granulation was identified as a suitable technique to treat this kind of wastewater due to excellent settleability, high biomass retention and a good ability to handle high organic loads and seasonal fluctuations. However, energy generation from anaerobic digestion plays an important role, favouring anaerobic membrane bioreactor application, although it was observed to be sensitive to sudden load fluctuations, which led to a thorough pH control and alkali addition.

Short- and long-term effects of ammonium and nitrite on the Anammox process.

Autotrophic anaerobic ammonium oxidation (Anammox) is a biological process in which Planctomycete-type bacteria combine ammonium and nitrite to generate nitrogen gas. Both substrates can exert inhibitory effects on the process, causing the decrease of the specific activity of the biomass and the loss of the stable operation of reactors. The aim of the present work is to evaluate these effects in short- and long-term experiments. The short-term effects were carried out with two different types of Anammox biomass, biofilm on inorganic carriers and flocculent sludge. The effects of ammonium on both kinds of biomass were similar. A decrease of the Specific Anammox Activity (SAA) of 50% was observed at concentrations about 38 mg NH(3)-N·L(-1), while 100 mg NH(3)-N·L(-1) caused an inhibition of 80%. With regards to nitrite, the SAA was not affected at concentrations up to 6.6 µg HNO(2)-N·L(-1) but it suffered a decrease over 50% in the presence of 11 µg HNO(2)-N·L(-1) in the case of the biofilm. The flocculent biomass was much less resistant and its SAA sharply decreased up to 30% of its initial value in the presence of 4.4 µg HNO(2)-N·L(-1). The study of the long-term effects was carried out in lab-scale Sequencing Batch Reactors (SBR) inoculated with the biofilm biomass. Concentrations up to 20 mg NH(3)-N·L(-1) showed no effects on either reactor efficiency or biomass activity. However, when free ammonia concentrations reached values between 35 and 40 mg NH(3)-N·L(-1), the operation turned unstable and the efficiency was totally lost. Nitrous acid concentrations around 1.5 µg HNO(2)-N·L(-1) caused a loss of the efficiency of the treatment and a destabilization of the system. However, a total restoration of the SAA was observed after the stoichiometric feeding was applied to the SBR.

Partial nitrification of sludge reject water by means of aerobic granulation.

Granular sludge formation was performed in a laboratory scale Sequencing Batch Reactor (SBR) fed with supernatant of anaerobic digestion of sewage sludge. This effluent was concentrated progressively in order to enhance biomass capacity without inhibiting it. During the first part of the study, ammonium nitrogen was converted to nitrate, so conventional nitrification took place. When a nitrogen load of 0.8 g N L(-1) d(-1) was treated, the effluent concentration of nitrite started to increase while the nitrate concentration decreased until it disappeared. So, partial nitrification was achieved. At the end of this study, a nitrogen load of 1.1 g N L(-1) d(-1) was treated obtaining an effluent with 50% ammonium and 50% nitrite. The volatile suspended solids concentration in the reactor reached 10 g VSS L(-1). At this point the granule morphology was quite round and no filamentous bacteria was observed. The Feret's diameter was in the range between 1 and 6 mm with an average value of 4.5 mm. Roundness value was all the time higher than 0.7. Granule density increased during the experimental period, obtaining a final value of 7.0 g L(-1).

Phenol removal from hypersaline wastewaters in a Membrane Biological Reactor (MBR): operation and microbiological characterisation.

In this study, two Membrane Biological Reactors (MBR) with submerged flat membranes, one at lab-scale conditions and the other at pilot-plant conditions, were operated at environmental temperature to treat an industrial wastewater characterised by low phenol concentrations (8-16 mg L(-1)) and high salinity (∼ 150-160 mS cm(-1)). During the operation of both reactors, the phenol loading rate was progressively increased and less than 1mg phenol L(-1) was detected even at very low HRTs (0.5-0.7 days). Membrane fouling was minimized by the cross flow aeration rate inside the MBRs and by intermittent permeation. Microbial community analysis of both reactors revealed that members of the genera Halomonas and Marinobacter (gammaproteobacteria) were major components. Growth-linked phenol degradation by pure cultures of Marinobacter isolates demonstrated that this bacterium played a major role in the removal of phenol from the bioreactors.

Start-up of an aerobic granular sequencing batch reactor for the treatment of winery wastewater.

Aerobic granular sludge was cultivated in a sequencing batch reactor (SBR) in order to remove the organic matter present in winery wastewater. The formation of granules was performed using a synthetic substrate. The selection parameter was the settling time, as well as the alternation of feast-famine periods, the air velocity and the height/diameter ratio of the reactor. After 10 days of operation under these conditions, the first aggregates could be observed. Filamentous bacteria were still present in the reactor but they disappeared progressively. During the start-up, COD loading was increased from 2.7 to 22.5 kg COD/(m(3) day) in order to obtain a feast period between 30 and 60 minutes. At this point, granules were quite round, with a particle diameter between 3.0 and 4.0 mm and an average density of 6 g L(-1). After 120 days of operation, synthetic media was replaced by real winery wastewater, with a COD loading of 6 kg COD/(m(3) day). The decrease of the organic load implied a reduction of the aggregate diameter and a density increase up to 13.2 g L(-1). The effluent was free of organic matter and the solids concentration in the reactor reached 6 g VSS L(-1).

SBR technology for high ammonium loading rates.

This paper focuses on the study of high ammonium concentrated wastewater with SBR reactors. Four type of wastewaters, landfill leachates (T=20 degrees C) and the reject water (T=35 degrees C) coming from mesophilic anaerobic digesters of sewage sludge, pig slurry and organic fraction of municipal solid waste (OFMSW), were studied in four SBR during 6 months. The removal of nitrogen was done in all the cases with nitrification/denitrification via nitrite obtaining high removal nitrogen conversions for the three types of reject water (0.75-0.85 kg N day(-1) m(-3)) and lower for landfill leachates due to temperature requirements (0.3 kg N day(-1) m(-3)).