Adaptive prediction for effluent quality of wastewater treatment plant: Improvement with a dual-stage attention-based LSTM network.

The accurate effluent prediction plays a crucial role in providing early warning for abnormal effluent and achieving the adjustment of feedforward control parameters during wastewater treatment. This study applied a dual-staged attention mechanism based on long short-term memory network (DA-LSTM) to improve the accuracy of effluent quality prediction. The results showed that input attention (IA) and temporal attention (TA) significantly enhanced the prediction performance of LSTM. Specially, IA could adaptively adjust feature weights to enhance the robustness against input noise, with R2 increased by 13.18%. To promote its long-term memory ability, TA was used to increase the memory span from 96 h to 168 h. Compared to a single LSTM model, the DA-LSTM model showed an improvement in prediction accuracy by 5.10%, 2.11%, 14.47% for COD, TP, and TN. Additionally, DA-LSTM demonstrated excellent generalization performance in new scenarios, with the R2 values for COD, TP, and TN increasing by 22.67%, 20.06%, and 17.14% respectively, while the MAPE values decreased by 56.46%, 63.08%, and 42.79%. In conclusion, the DA-LSTM model demonstrated excellent prediction performance and generalization ability due to its advantages of feature-adaptive weighting and long-term memory focusing. This has forward-looking significance for achieving efficient early warning of abnormal operating conditions and timely management of control parameters.

Electroactive biocake layer-driven advanced removal of dissolved organic matter at membrane interface of anaerobic electrochemical membrane bioreactor.

The bio-cake layer is one of the most negative effects during water and wastewater filtration, but its potential behoof of biodegradation is poorly understood. In this study, we activated and reconstructed the bio-cake by using the carbon nanotube membrane (25 cm2 area, 17 LMH flux) as the anode in an anaerobic membrane bioreactor (AnMBR), and investigated its positive role in advanced removal of dissolved organic matter from up-flow anaerobic sludge bed unit (3 L/d) when treating synthetic municipal wastewater. At the anodic membrane interface, the enhanced biodegradation was proved to dominate the DOM reduction (contribution >40%), controlling the effluent COD as low as 19.2 ± 2.5 mg/L. Bio-cake characterizations suggested that the positive potential induced electroactive improvement, cell viability boost, and metabolic optimization. Metatranscriptomic analyses revealed that anode respiratory out-compete methanogenesis, forwarding a synergetic metabolism between enriched fermenters like Proteiniphilum sp. and exoelectrogens like Geobacter sp. Thus, electroactive bio-cake not only accelerated the decomposition of inside foulants to maintain the high flux, but also efficiently intercepted flow-through DOM due to reduced mass-transfer limitations and enhanced metabolic activity. An ordered, non-clogging, and potentially functional "cell filter" was established to achieve a win-win situation between fouling control and effluent improvement, which is promising to upgrade the AnMBR technology for maximizing the sustainable regeneration in future wastewater treatment.

Biological performance and fouling mitigation in the biochar-amended anaerobic membrane bioreactor (AnMBR) treating pharmaceutical wastewater.

Anaerobic membrane bioreactor (AnMBR) is an advanced technology in treating pharmaceutical wastewater, but the membrane fouling limits its development. In this study, the biochar with adsorption capacity of biopolymers was added in AnMBR to investigate its potential in treating pharmaceutical wastewater and alleviating membrane fouling. In the biochar-amended AnMBR, adsorbable organic halogen (AOX) was removed effectively, and more COD was biotransformed into CH4. Membrane fouling mitigation was achieved in the third stage with a 56% decrease of average transmembrane pressure difference (TMP) rising rate. The predominant culprit, proteins of extracellular polymeric substance (EPS-proteins) in sludge mixture and cake layer, was reduced significantly. Particularly, the proportion of micromolecular (0.1-0.15 kDa) EPS-proteins in cake layer was 1.5-folds that of the control group. The important bio-foulant genus Arcobacter aggregating on the membrane had less and almost half the relative abundance (16.5%) than that of the control group (30.7%).