Neural network modeling for on-line estimation of nutrient dynamics in a sequentially-operated batch reactor.

In monitoring and controlling wastewater treatment processes, on-line information of nutrient dynamics is very important. However, these variables are determined with a significant time delay. Although the final effluent quality can be analyzed after this delay, it is often too late to make proper adjustments. In this paper, a neural network approach, a software sensor, was proposed to overcome this problem. Software sensor refers to a modeling approach inferring hard-to-measure process variables from other on-line measurable process variables. A bench-scale sequentially-operated batch reactor (SBR) used for advanced wastewater treatment (BOD plus nutrient removal) was employed to develop the neural network model. In order to improve the network performance, the structure of neural network was arranged in such a way of reflecting the change of operational conditions within a cycle. Real-time estimation of PO3-(4), NO-3, and NH+4 concentrations was successfully carried out with the on-line information of the SBR system only.     

Control of carbon and ammonium ratio for simultaneous nitrification and denitrification in a sequencing batch bioreactor

This study shows how the carbon and nitrogen (C/N) ratio controls the simultaneous occurrence of nitrification and denitrification in a sequencing batch reactor (SBR). Data demonstrated that a low C/N ratio resulted in a rapid carbon deficit, causing an unbalanced simultaneous nitrification–denitrification (SND) process in SBR. When the initial COD/NH₄⁺-N ratio was adjusted to 11.1, the SND-based SBR achieved complete removal of NH₄-N and COD without leaving any NO₂⁻-N in the effluent. The nitrogen removal efficiency decreases gradually with increasing ammonium-loading rate to the SND–SBR system. Altogether, data showed that appropriate controls of carbon and nitrogen input are required to achieve an efficient SND–SBR. An established SND technology can save operation time and energy, and might replace the traditional two-stage biological nitrification and denitrification process.     

A sequencing batch reactor system for high-level biological nitrogen and phosphorus removal from abattoir wastewater

A sequencing batch reactor (SBR) system is demonstrated to biologically remove nitrogen, phosphorus and chemical oxygen demand (COD) to very low levels from abattoir wastewater. Each 6 h cycle contained three anoxic/anaerobic and aerobic sub-cycles with wastewater fed at the beginning of each anoxic/anaerobic period. The step-feed strategy was applied to avoid high-level build-up of nitrate or nitrite during nitrification, and therefore to facilitate the creation of anaerobic conditions required for biological phosphorus removal. A high degree removal of total phosphorus (>98%), total nitrogen (>97%) and total COD (>95%) was consistently and reliably achieved after a 3-month start-up period. The concentrations of total phosphate and inorganic nitrogen in the effluent were consistently lower than 0.2 mg P l⁻¹ and 8 mg N l⁻¹, respectively. Fluorescence in situ hybridization revealed that the sludge was enriched in Accumulibacter spp. (20–40%), a known polyphosphate accumulating organism, whereas the known glycogen accumulating organisms were almost absent. The SBR received two streams of abattoir wastewater, namely the effluent from a full-scale anaerobic pond (75%) and the effluent from a lab-scale high-rate pre-fermentor (25%), both receiving raw abattoir wastewater as feed. The pond effluent contained approximately 250 mg N l⁻¹ total nitrogen and 40 mg P l⁻¹ of total phosphorus, but relatively low levels of soluble COD (around 500 mg l⁻¹). The high-rate lab-scale pre-fermentor, operated at 37°C and with a sludge retention time of 1 day, proved to be a cheap and effective method for providing supplementary volatile fatty acids allowing for high-degree of biological nutrient removal from abattoir wastewater.     

Effects of ferrous iron on the performance and microbial community in aerobic granular sludge in relation to nutrient removal

Lab‐scale experiments were conducted to investigate the effects of ferrous iron on nutrient removal performance and variations in the microbial community inside aerobic granular sludge for 408 days. Two reactors were simultaneously operated, one without added ferrous iron (SBR1), and one with 10 mg Fe²⁺ L^?1 of added ferrous iron (SBR2). A total of 1 mg Fe²⁺ L^?1 of added ferrous iron was applied to SBR1 starting from the 191st day to observe the resulting variations in the nutrient removal performance and the microbial community. The results show that ammonia‐oxidizing bacteria (AOB) could not oxidize ammonia due to a lack of iron compounds, but they could survive in the aerobic granular sludge. Limited ferrous iron addition encouraged nitrification. Enhanced biological phosphorus removal (EBPR) from both reactors could not be maintained regardless of the amount of ferrous iron that was applied. EBPR was established in both reactors when the concentration of mixed liquor suspended solid (MLSS) and the percentage of Accumulibacteria increased. A total of 10 mg Fe²⁺ L^?1 of added ferrous iron had a relatively adverse effect on the growth of AOB species compared to 1 mg Fe²⁺ L^?1 of added ferrous iron, but it encouraged the growth of Nitrospira sp. and Accumulibacteria, which requires further study. It could be said that the compact and stable structure of aerobic granular sludge preserved AOB and NOB from Fe‐deficient conditions, and wash‐out during the disintegration period. © 2017 American Institute of Chemical Engineers Biotechnol. Prog., 33:716–725, 2017     

Simulating a cyclic activated sludge system by employing a modified ASM3 model for wastewater treatment

To interpret the biological nutrient removal in a cyclic activated sludge system (CAS), a modified model was developed by combining the process of simultaneous storage and growth, and the kinetics of soluble microbial product (S _SMP) and extracellular polymeric substance (X _EPS) with activated sludge model no. 3 (ASM3). These most sensitive parameters were initially selected whilst parameters with low sensitivity were given values from literature. The selected parameters were then calibrated on an oxygen uptake rate test and a batch CAS reactor on an operational cycle. The calibrated model was validated using a combination of the measurements from a batch CAS reactor operated for 1 month and the average deviation method. The simulations demonstrated that the modified model was capable of predicting higher effluent concentrations compared to outputs of the ASM3 model. Additionally, it was also shown that the average deviation of effluent S _COD, S _NH, S _SMP and X _EPS simulated with the modified model was all less than 1 mg L⁻¹. In summary, the model could effectively describe biological processes in a CAS reactor and provide a wonderful tool for operation.

     

Use of Stable-Isotope Probing, Full-Cycle rRNA Analysis, and Fluorescence In Situ Hybridization-Microautoradiography To Study a Methanol-Fed Denitrifying Microbial Community

A denitrifying microbial consortium was enriched in an anoxically operated, methanol-fed sequencing batch reactor (SBR) fed with a mineral salts medium containing methanol as the sole carbon source and nitrate as the electron acceptor. The SBR was inoculated with sludge from a biological nutrient removal activated sludge plant exhibiting good denitrification. The SBR denitrification rate improved from less than 0.02 mg of NO₃⁻-N mg of mixed-liquor volatile suspended solids (MLVSS)⁻¹ h⁻¹ to a steady-state value of 0.06 mg of NO₃⁻-N mg of MLVSS⁻¹ h⁻¹ over a 7-month operational period. At this time, the enriched microbial community was subjected to stable-isotope probing (SIP) with [¹³C]methanol to biomark the DNA of the denitrifiers. The extracted [¹³C]DNA and [¹²C]DNA from the SIP experiment were separately subjected to full-cycle rRNA analysis. The dominant 16S rRNA gene phylotype (group A clones) in the [¹³C]DNA clone library was closely related to those of the obligate methylotrophs Methylobacillus and Methylophilus in the order Methylophilales of the Betaproteobacteria (96 to 97% sequence identities), while the most abundant clone groups in the [¹²C]DNA clone library mostly belonged to the family Saprospiraceae in the Bacteroidetes phylum. Oligonucleotide probes for use in fluorescence in situ hybridization (FISH) were designed to specifically target the group A clones and Methylophilales (probes DEN67 and MET1216, respectively) and the Saprospiraceae clones (probe SAP553). Application of these probes to the SBR biomass over the enrichment period demonstrated a strong correlation between the level of SBR denitrification and relative abundance of DEN67-targeted bacteria in the SBR community. By contrast, there was no correlation between the denitrification rate and the relative abundances of the well-known denitrifying genera Hyphomicrobium and Paracoccus or the Saprospiraceae clones visualized by FISH in the SBR biomass. FISH combined with microautoradiography independently confirmed that the DEN67-targeted cells were the dominant bacterial group capable of anoxic [¹⁴C]methanol uptake in the enriched biomass. The well-known denitrification lag period in the methanol-fed SBR was shown to coincide with a lag phase in growth of the DEN67-targeted denitrifying population. We conclude that Methylophilales bacteria are the dominant denitrifiers in our SBR system and likely are important denitrifiers in full-scale methanol-fed denitrifying sludges.     

Effects of glucose and phenol on soluble microbial products (SMP) in sequencing batch reactor systems

Soluble microbial products (SMP) are ubiquitously present in the effluents of biological wastewater treatment systems. In sequencing batch reactor (SBR) systems, effects of influent concentration and temperature on the amount and the molecular weight (MW) distribution of SMP were investigated for the two substrates, glucose and phenol. The values of effluent SMP/S₀ of phenol were higher than those of glucose at different influent concentrations and temperatures. It was found that the effluent SMP (S_e) was linearly correlated to the influent total organic carbon (TOC) (S₀) for both substrates. The slope and intercept of the equation were affected by the temperature. According to the analysis of the MW distribution, it was shown that there exists a bimodal pattern with the majority of SMP having a MW<1 kDa or >10 kDa. The low MW fraction (<1 kDa) amounts to 47.3–70.4% of the effluent SMP. The high MW fraction (>10 kDa) slightly fluctuates in the range of 21.2–32.8% of the effluent SMP.     

Biologically Fe<Superscript>2+</Superscript> oxidizing fluidized bed reactor performance and controlling of Fe<Superscript>3+</Superscript> recycle during heap bioleaching: an artificial neural network-based model

The performance of a biological Fe²⁺ oxidizing fluidized bed reactor (FBR) was modeled by a popular neural network-back-propagation algorithm over a period of 220 days at 37 °C under different operational conditions. A method is proposed for modeling Fe³⁺ production in FBR and thereby managing the regeneration of Fe³⁺ for heap leaching application, based on an artificial neural network-back-propagation algorithm. Depending on output value, relevant control strategies and actions are activated, and Fe³⁺ production in FBR was considered as a critical output parameter. The modeling of effluent Fe³⁺ concentration was very successful, and an excellent match was obtained between the measured and the predicted concentrations.     

Coupled biological and photo-Fenton pretreatment system for the removal of di-(2-ethylhexyl) phthalate (DEHP) from water

The photo-Fenton coupled with a biological system for the removal of di-(2-ethylhexyl) phthalate (DEHP) in wastewater was analyzed. The toxicity of DEHP-containing wastewater was found to be reduced after pretreatment by the photo-Fenton reaction. The effect of different factors, such as DEHP, Fe³⁺ and H₂O₂ concentrations and the reaction time, on degradation efficiency was investigated. The optimal time to stop the pretreatment process and introduce the effluent to the biological system was 60 min. The results show that effluent of DEHP-containing wastewater pretreated by the photo-Fenton method is biodegradable and that mineralization can be completed when the wastewater is subsequently treated in a biological system. The coupled Fenton and biological treatment system for the degradation of DEHP-containing wastewater can be successfully performed in a semi-continuous mode. These results indicate that the coupled photo-biological system is an effective and potential method for the treatment of DEHP-containing wastewater.     

Optimization of actinomycin V production by Streptomyces triostinicus using artificial neural network and genetic algorithm

Artificial neural network (ANN) and genetic algorithm (GA) were applied to optimize the medium components for the production of actinomycinV from a newly isolated strain of Streptomyces triostinicus which is not reported to produce this class of antibiotics. Experiments were conducted using the central composite design (CCD), and the data generated was used to build an artificial neural network model. The concentrations of five medium components (MgSO₄, NaCl, glucose, soybean meal and CaCO₃) served as inputs to the neural network model, and the antibiotic yield served as outputs of the model. Using the genetic algorithm, the input space of the neural network model was optimized to find out the optimum values for maximum antibiotic yield. Maximum antibiotic yield of 452.0 mg l⁻¹ was obtained at the GA-optimized concentrations of medium components (MgSO₄ 3.657; NaCl 1.9012; glucose 8.836; soybean meal 20.1976 and CaCO₃ 13.0842 gl⁻¹). The antibiotic yield obtained by the ANN/GA was 36.7% higher than the yield obtained with the response surface methodology (RSM).     

Sewage effluent discharge and geothermal input in a natural wetland,Tongariro Delta,New Zealand

Effluent from the oxidation ponds of the town of Turangi, south of Lake Taupo, has been discharged into a natural wetland since the 1960s. This has resulted in elevated concentrations of Na⁺, K⁺, Cl⁻ and NH₄⁺-N in both ground and surface water. Increased weed invasion and plant growth, and high heavy metal concentrations (e.g. up to 440 ppm Zn) occur in the vicinity of effluent discharge in the wetland. Element and nutrient concentrations decrease with increasing distance from the effluent inflow point, suggesting that the wetland presently acts as a sink for metals and nutrients. Elevated arsenic concentrations (up to 5800 ppm in peat and 11 400 ppm in the ash fraction) in some of the organic-rich sediment suggest a long-term input by geothermal water originating in the Tokaanu–Waihi field. Increased silt input due to more frequent flooding of the Tongariro river over the past 40 years has resulted in a significant change in stratigraphy (from peat to mud) over much of the wetland.     

Neural network modeling for nutrient dynamics in a recycling piggery slurry treatment system

A recycling reactor system operated under sequential anoxic and oxic conditions was evaluated, in which the nutrients of piggery slurry were anaerobically and aerobically treated and then a portion of the effluent was recycled to the pigsty. The most dominant aerobic heterotrophs from the reactor were Alcaligenes faecalis (TSA-3), Brevundimonas diminuta (TSA-1) and Abiotrophia defectiva (TSA-2) in decreasing order, whereas lactic acid bacteria, LAB (MRS-1, etc.) were most dominantly observed in the anoxic tank. Here we have tried to model the nutrient removal process for each tank in the system based on population densities of heterotrophic and LAB. Principal component analysis (PCA) was first applied to delineate a relationship between input (microbial densities and treatment parameters such as population densities of heterotrophic and LAB, suspended solids (SS), COD, NH₄ ⁺–N, ortho-phosphorus, and total phosphorus) and output. Multi-layer neural networks using an error back-propagation learning algorithm were then employed to model the nutrient removal process for each tank. PCA filtration of microbial densities as input data was able to enhance generalization performance of the neural network, and this has led to a better prediction of the measured data. Neural networks independently trained for each treatment tank and the combined analysis of the subsequent tank data allowed a successful prediction of the treatment system for at least 2 days.     

Impacts of the Reduction of Nutrient Levels on Bacterial Water Quality in Distribution Systems

This study evaluated the impacts of reducing nutrient levels on bacterial water quality in drinking water. Two American Water System facilities (sites NJ102a and IN610) with histories of coliform problems were involved, and each water utility received two pilot distribution systems (annular reactors). One reactor simulated the conventional treatment conditions (control), while the other reactor was used to assess the effect of biological filtration and subsequent reduced biodegradable organic matter levels on suspended (water column) and biofilm bacterial concentrations in the distribution systems. Biodegradable organic matter levels were reduced approximately by half after biological treatment. For site NJ102a, the geometric mean of the assimilable organic carbon concentrations was 217 μg/liter in the plant effluent and 91 μg/liter after biological filtration. For both sites, plant effluent biodegradable dissolved organic carbon levels averaged 0.45 mg/liter, versus 0.19 to 0.22 mg/liter following biological treatment. Biological treatment improved the stability of free chlorine residuals, while it had little effect on chloramine consumption patterns. High bacterial levels from the biological filters resulted in higher bacterial concentrations entering the test reactors than entering the control reactors. On average, biofilms in the model systems were reduced by 1 log unit (from 1.4 × 10⁵ to 1.4 × 10⁴ CFU/cm²) and 0.5-log unit (from 2.7 × 10⁵ to 7.8 × 10⁴ CFU/cm²) by biological treatment at sites NJ102a and IN610, respectively. Interestingly, it required several months of biological treatment before there was an observable impact on bacterial water quality in the system, suggesting that the effect of the treatment change was influenced by other factors (i.e., pipe conditions or disinfection, etc.).     

Growth ofPhormidium sp. in aerobic secondary piggery waste-water

The potential of a cyanobacteriumPhormidium sp., for the tertiary treatment of piggery waste-water chemical oxygen demand , 3000 mg·1^–1, using an aerobically stabilized secondary effluent, was studied. Batch cultures were carried out in 2-1 glass reactors and in a 30-1 glass-fibre carousel reactor. The nutrient removal efficiency as well as biomass production were compared in synthetic mineral medium and in different concentrations of aerobically stabilized piggery effluent. The best performance ofPhormidium sp. ocurred in diluted stabilized secondary effluent (1:1). Removal efficiencies were 100% for P-PO₄ ^–3, 50% for N-NH₄ ^–1 and 35% for N-NO₃ ^– in small-volume cultures; and 31% for P-PO₄ ^–3, 100% for N-NH₄ ^– and 70% for N-NO₃ ^– in an open carousel reactor. Biomass production on the aerobically stabilized effluent was six times higher in the 2-1 reactors and 1.7 times higher in the carousel reactor when compared to the synthetic medium.     

Regulation of nitrogen levels for heterogeneous populations of sewage origin grown in completely mixed reactors

Heterogeneous populations of sewage origin were grown continuously at, dilution rates from 1/12 hr^?1 to dilute-out (1/1 hr^?1) using glucose (1000 mg/l) as carbon source and three concentrations of NH₃-N as the nitrogen source (COD:N = 70:1, 40:1, and 25:1). The effects of nitrogen level and growth rate (dilution rate) on substrate removal, biological solids production, cellular carbohydrate and protein, and NH-N in the effluent were examined. It was found that the optimum level of nitrogen supplementation for the synthetic nitrogen-deficient waste employed should not be based solely on the desired effluent quality with respect to COD removal but should include due consideration of reactor detention time (or dilution rate) and the allowable (or desirable) level of nitrogen leakage in the effluent.     

Bioremediation of Perchlorate-Contaminated Groundwater Using a Packed Bed Biological Reactor

The objective of this study was to assess the efficacy of a bench-scale, acetate-fed, packed bed bioreactor (PBR) to treat low concentrations (>1 mg L^?1) of perchlorate (ClO₄ ^?) in groundwater collected from an impacted site. The PBR consisted of a cylindrical plexiglass column packed with Celite, a diatomaceous earth product, as a solid support medium. The reactor was inoculated with a ClO_?4 ^?-reducing bacterial isolate, perclace. Results showed that with influent ClO₄ ^? concentrations of approximately 800 μg L^?1, nondetectable effluent concentrations (>4 μg L^?1) were achieved with the PBR/perclace system at residence time as low as 0.3 h. Influent acetate concentrations of less than 500 mg L^?1 yielded nondetectable effluent ClO^? ₄ concentrations, and acetate concentrations generally less than 50 mg L^?1 were present in the effluent. Nitrate (NO^? _?3) was also removed in this system, while sulfate (SO₄ ^2?) reduction was not observed. The pH remained relatively constant during the process.     

How do chronic nutrient loading and the duration of nutrient pulses affect nutrient uptake in headwater streams?

Our study aimed to analyze the effects of chronic nutrient loading on the capacity of headwater streams to retain phosphorus and ammonium pulses of different duration. For this purpose, we selected nine headwater streams located across a gradient of increasing agricultural land use and eutrophication. In each stream, we performed sequential plateau additions with increasing nutrient concentrations in summer 2015 and instantaneous slug additions in summer 2016 under similar hydrological conditions. We modelled kinetic uptake curves from the slug additions via the Tracer Additions for Spiraling Curve Characterization method and calculated ambient uptake parameters. Ambient uptake rates generally increased (1.4–20.8 µg m^?2 s^?1 for NH₄–N and 0.3–10.3 µg m^?2 s^?1 for SRP, respectively), while ambient uptake velocities decreased from oligotrophic to polytrophic streams (1.8–14.0 mm min^?1 for NH₄–N and 1.6–9.9 mm min^?1 for SRP, respectively). However, correlations between ambient uptake parameters and background concentrations were weak. Concentration-dependent uptake rates followed either a linear or a Michaelis–Menten saturation model, regardless of the degree of nutrient loading. Uptake rate curves showed counter-clockwise hysteresis in oligotrophic streams and clockwise hysteresis in streams of higher trophic states, indicating a reduced significance of hyporheic uptake with increasing nutrient loading. Comparisons of slug and plateau additions revealed that oligotrophic streams were most efficient in uptake during short nutrient pulses, while eutrophic streams profited from longer pulse duration. The results indicate that nutrient uptake is increasingly transport-controlled in polluted streams where increased biofilm thickness and clogging of sediments restrict nutrient transport to reactive sites.     

Vertical pattern and its driving factors in soil extracellular enzyme activity and stoichiometry along mountain grassland belts

Headwater streams are foci for nutrient and energy loading from terrestrial landscapes, in situ nutrient transformations, and downstream transport. Despite the prominent role that headwater streams can have in regulating downstream water quality, the relative importance of processes that can influence nutrient uptake have not been fully compared in heterotrophic aquatic systems. To address this research need, we assessed the seasonality of dissolved organic carbon (DOC) and nitrate (NO₃^?) uptake, compared the relative influence of hydrologic and biogeochemical drivers on observed seasonal trends in nutrient uptake, and estimated the influence of these biological transformations on watershed scale nutrient retention and export. We determined that seasonal reductions in DOC and NO₃^? concentrations led to decreases in the potential for the biotic community to take up nutrients, and that seasonality of DOC and NO₃^? uptake was consistent with the seasonal dynamics of ecosystem metabolism. We calculated that that during the post-snowmelt period (June to August), biotic retention of both dissolved organic carbon and nitrate exceeded export fluxes from this headwater catchment, highlighting the potential for biological processes to regulate downstream water quality.     

Aerobic granulation with low strength wastewater at low aeration rate in A/O/A SBR reactor

The formation of aerobic granules with low organic loading synthetic wastewater (150-200 mg L⁻¹ of influent COD, acetate/propionate = 1/3) at low aeration rate (0.6 cm s⁻¹ of superficial gas velocity) had been investigated in the anaerobic/oxic/anoxic SBR. Aerobic granules with smooth surface and compact structure were successfully obtained after 50 days. However, these aerobic granules were unstable when the d(0.9) of granules increased to more than 1 mm. The results suggested that the aerobic granules with small diameter (smaller than 1000 μm) were more favorable for treating the low substrate loading wastewater at the low aeration rate. The cycle test revealed that most of the influent COD was removed at the anaerobic stage. The effluent concentrations of N-NH₄⁺ and P-PO₄³⁻ were lower than 1 mg L⁻¹, and the effluent concentration of nitrate gradually decreased with the granulation. Phosphate accumulating organisms were found to utilize O₂ or NO_x⁻ as electron acceptor for phosphorus removal in the study. Simultaneous nitrogen and phosphorus removal occurred inside the granules.