Machine learning enabled prediction and process optimization of VFA production from riboflavin-mediated sludge fermentation

Highlights

● Data-driven approach was used to simulate VFA production from WAS fermentation.

● Three machine learning models were developed and evaluated.

● XGBoost showed best prediction performance and excellent generalization ability.

● pH and protein were the top two input features for the modeling.

● The maximal VFA production was predicted to be 650 mg COD/g VSS.

Abstract

Riboflavin is a redox mediator that promotes volatile fatty acids (VFAs) production from waste activated sludge (WAS) and is a promising method for WAS reuse. However, time- and labor-consuming experiments challenge obtaining optimal operating conditions for maximal VFA production. In this study, three machine learning (ML) models were developed to predict the VFAs production from riboflavin-mediated WAS fermentation systems. Among the three tested ML algorithms, eXtreme Gradient Boosting (XGBoost) presented the best prediction performance and excellent generalization ability, with the highest testing coefficient of determination (R² of 0.93) and lowest root mean square error (RMSE of 0.070). Feature importance analysis and their interactions using the Shepley Additive Explanations (SHAP) method indicated that pH and soluble protein were the top two input features for the modeling. The intrinsic correlations between input features and microbial communities corroborated this deduction. On the optimized ML model, genetic algorithm (GA) and particle swarm optimization (PSO) solved the optimal solution of VFA output, predicting the maximum VFA output as 650 mg COD/g VSS. This study provided a data-driven approach to predict and optimize VFA production from riboflavin-mediated WAS fermentation.

Graphical abstract

Keywords

Machine learning / Volatile fatty acids / Riboflavin / Waste activated sludge / eXtreme Gradient Boosting

References

[1] Arudchelvam Y , Perinpanayagam M , Nirmalakhandan N . (2010). Predicting VFA formation by dark fermentation of particulate substrates. Bioresource Technology, 101(19): 7492–7499

[2] Battista F , Strazzera G , Valentino F , Gottardo M , Villano M , Matos M , Silva F , Reis M A , Mata-Alvarez J , Astals S . (2022). New insights in food waste, sewage sludge and green waste anaerobic fermentation for short-chain volatile fatty acids production: a review. Journal of Environmental Chemical Engineering, 10(5): 108319

[3] Bevilacqua R , Regueira A , Mauricio-Iglesias M , Lema J M , Carballa M . (2020). Protein composition determines the preferential consumption of amino acids during anaerobic mixed-culture fermentation. Water Research, 183: 115958

[4] Bhandari V , Gupta R S . (2012). Molecular signatures for the phylum Synergistetes and some of its subclades. Antonie van Leeuwenhoek, 102(4): 517–540

[5] Breiman L . (2001). Random forests. Machine Learning, 45(1): 5–32

[6] Chen T, Guestrin C (2016). XGBoost: a scalable tree boosting system. San Francisco, California, USA: Association for Computing Machinery, 785–794

[7] Chen Y , Jiang S , Yuan H , Zhou Q , Gu G . (2007). Hydrolysis and acidification of waste activated sludge at different pHs. Water Research, 41(3): 683–689

[8] Chen Y , Jiang X , Xiao K , Shen N , Zeng R , Zhou Y . (2017). Enhanced volatile fatty acids (VFAs) production in a thermophilic fermenter with stepwise pH increase–Investigation on dissolved organic matter transformation and microbial community shift. Water Research, 112: 261–268

[9] Andrade Cruz I , Chuenchart W , Long F , Surendra K C , Renata Santos Andrade L , Bilal M , Liu H , Tavares Figueiredo R , Khanal S K , Fernando Romanholo Ferreira L . (2022). Application of machine learning in anaerobic digestion: perspectives and challenges. Bioresource Technology, 345: 126433

[10] D’Silva T C , Isha A , Chandra R , Vijay V K , Subbarao P M V , Kumar R , Chaudhary V P , Singh H , Khan A A , Tyagi V K . . (2021). Enhancing methane production in anaerobic digestion through hydrogen assisted pathways: a state-of-the-art review. Renewable & Sustainable Energy Reviews, 151: 111536

[11] De Clercq D , Wen Z , Fei F , Caicedo L , Yuan K , Shang R . (2020). Interpretable machine learning for predicting biomethane production in industrial-scale anaerobic co-digestion. Science of the Total Environment, 712: 134574

[12] Du Y , Wang M , Yang L , Tong L , Guo D , Ji X . (2022). Optimization and scale-up of fermentation processes driven by models. Bioengineering (Basel, Switzerland), 9(9): 473

[13] Fang W , Zhang X , Zhang P , Wan J , Guo H , Ghasimi D S , Morera X C , Zhang T . (2020). Overview of key operation factors and strategies for improving fermentative volatile fatty acid production and product regulation from sewage sludge. Journal of Environmental Sciences (China), 87: 93–111

[14] Gahlot P , Balasundaram G , Tyagi V K , Atabani A , Suthar S , Kazmi A , Štěpanec L , Juchelková D , Kumar A . (2022). Principles and potential of thermal hydrolysis of sewage sludge to enhance anaerobic digestion. Environmental Research, 214(2): 113856

[15] Guo Q , Yang Z , Zhang B , Hua M , Liu C , Pan B . (2022). Enhanced methane production during long-term UASB operation at high organic loads as enabled by the immobilized Fungi. Frontiers of Environmental Science & Engineering, 16(6): 71

[16] Hecht J S , Zia A , Clemins P J , Schroth A W , Winter J M , Oikonomou P D , Rizzo D M . (2022). Modeling the sensitivity of cyanobacteria blooms to plausible changes in precipitation and air temperature variability. Science of the Total Environment, 812: 151586

[17] Hosseinzadeh A , Zhou J L , Altaee A , Li D . (2022). Machine learning modeling and analysis of biohydrogen production from wastewater by dark fermentation process. Bioresource Technology, 343: 126111

[18] Huang J , Chen S , Wu W , Chen H , Guo K , Tang J , Li J . (2019). Insights into redox mediator supplementation on enhanced volatile fatty acids production from waste activated sludge. Environmental Science and Pollution Research International, 26(26): 27052–27062

[19] Kazemi P , Bengoa C , Steyer J P , Giralt J . (2021). Data-driven techniques for fault detection in anaerobic digestion process. Process Safety and Environmental Protection, 146: 905–915

[20] Kazemi P , Steyer J P , Bengoa C , Font J , Giralt J . (2020). Robust data-driven soft sensors for online monitoring of volatile fatty acids in anaerobic digestion processes. Processes (Basel, Switzerland), 8(1): 67

[21] Kennedy J , Eberhart R . (1995). Particle swarm optimization. In: Proceedings of ICNN’95-international conference on neural networks. IEEE, 4: 1942–1948

[22] LeCun Y , Bengio Y , Hinton G . (2015). Deep learning. Nature, 521(7553): 436–444

[23] Li J , Pan L , Suvarna M , Wang X . (2021). Machine learning aided supercritical water gasification for H₂-rich syngas production with process optimization and catalyst screening. Chemical Engineering Journal, 426: 131285

[24] Li J , Zhang L , Li C , Tian H , Ning J , Zhang J , Tong Y W , Wang X . (2022). Data-driven based in-depth interpretation and inverse design of anaerobic digestion for CH₄-rich biogas production. ACS ES&T Engineering, 2(4): 642–652

[25] Liang T , Elmaadawy K , Liu B , Hu J , Hou H , Yang J . (2021). Anaerobic fermentation of waste activated sludge for volatile fatty acid production: recent updates of pretreatment methods and the potential effect of humic and nutrients substances. Process Safety and Environmental Protection, 145: 321–339

[26] Liu H , Han P , Liu H , Zhou G , Fu B , Zheng Z . (2018). Full-scale production of VFAs from sewage sludge by anaerobic alkaline fermentation to improve biological nutrients removal in domestic wastewater. Bioresource Technology, 260: 105–114

[27] Liu J , Huang J , Li H , Shi B , Xu Y , Liu J , Zhang D , Tang J , Hou P . (2022). Effect of temperature on fermentative VFAs production from waste sludge stimulated by riboflavin and the shifts of microbial community. Water Science and Technology, 85(4): 1191–1201

[28] Long F , Wang L , Cai W , Lesnik K , Liu H . (2021). Predicting the performance of anaerobic digestion using machine learning algorithms and genomic data. Water Research, 199: 117182

[29] Lu W , Huo W , Gulina H , Pan C . (2022). Development of machine learning multi-city model for municipal solid waste generation prediction. Frontiers of Environmental Science & Engineering, 16(9): 119

[30] Lundberg S M, Lee S I (2017). A unified approach to interpreting model predictions. International Conference on Neural Information Processing Systems,30: 4768–4777

[31] Luo J , Fang S , Huang W , Wang F , Zhang L , Fang F , Cao J , Wu Y . (2022). New insights into different surfactants’ impacts on sludge fermentation: focusing on the particular metabolic processes and microbial genetic traits. Frontiers of Environmental Science & Engineering, 16(8): 106

[32] Malhotra M , Garg A . (2021). Characterization of value-added chemicals derived from the thermal hydrolysis and wet oxidation of sewage sludge. Frontiers of Environmental Science & Engineering, 15(1): 13

[33] Pedregosa F , Varoquaux G , Gramfort A , Michel V , Thirion B , Grisel O , Blondel M , Prettenhofer P , Weiss R , Dubourg V . (2011). Scikit-learn: machine learning in Python. Journal of Machine Learning Research, 12: 2825–2830

[34] Piri J , Pirzadeh B , Keshtegar B , Givehchi M . (2021). Reliability analysis of pumping station for sewage network using hybrid neural networks-genetic algorithm and method of moment. Process Safety and Environmental Protection, 145: 39–51

[35] Qin S , Wainaina S , Liu H , Soufiani A M , Pandey A , Zhang Z , Awasthi M K , Taherzadeh M J . (2021). Microbial dynamics during anaerobic digestion of sewage sludge combined with food waste at high organic loading rates in immersed membrane bioreactors. Fuel, 303: 121276

[36] Ramsay I R , Pullammanappallil P C . (2001). Protein degradation during anaerobic wastewater treatment: derivation of stoichiometry. Biodegradation, 12(4): 247–257

[37] Regueira A , Lema J M , Carballa M , Mauricio-Iglesias M . (2020). Metabolic modeling for predicting VFA production from protein-rich substrates by mixed-culture fermentation. Biotechnology and Bioengineering, 117(1): 73–84

[38] Shehadeh A , Alshboul O , Al Mamlook R E , Hamedat O . (2021). Machine learning models for predicting the residual value of heavy construction equipment: an evaluation of modified decision tree, LightGBM, and XGBoost regression. Automation in Construction, 129: 103827

[39] Shi B , Huang J , Lin Y , Han W , Qiu S , Zhang D , Tang J , Hou P . (2023). Towards valeric acid production from riboflavin-assisted waste sludge: pH-dependent fermentation and microbial community. Waste and Biomass Valorization, 14: 833–845

[40] Shi B , Huang J , Yin Z , Han W , Qiu S , Tang J , Hou P . (2020). Riboflavin boosts fermentative valeric acid generation from waste activated sludge. BioResources, 15(2): 3962–3969

[41] Strazzera G , Battista F , Garcia N H , Frison N , Bolzonella D . (2018). Volatile fatty acids production from food wastes for biorefinery platforms: a review. Journal of Environmental Management, 226: 278–288

[42] Sun C , Xia A , Liao Q , Fu Q , Huang Y , Zhu X , Wei P , Lin R , Murphy J D . (2018). Improving production of volatile fatty acids and hydrogen from microalgae and rice residue: effects of physicochemical characteristics and mix ratios. Applied Energy, 230: 1082–1092

[43] Tang C , Yao X , Jin H , Sun Q , Zou Z , Yang W , He Z , Zhou A , Chen F , Ren Y . . (2022). Stepwise freezing-thawing treatment promotes short-chain fatty acids production from waste activated sludge. Science of the Total Environment, 818: 151694

[44] Vasudevan M, Natarajan N (2022). Towards achieving sustainable bioplastics production and nutrient recovery from wastewater: a comprehensive overview on polyhydroxybutyrate. Biomass Conversion and Biorefinery, DOI:10.1007/s13399-022-02399-z

[45] Vigneron A , Alsop E B , Lomans B P , Kyrpides N C , Head I M , Tsesmetzis N . (2017). Succession in the petroleum reservoir microbiome through an oil field production lifecycle. ISME Journal, 11(9): 2141–2154

[46] Wang J , Cao L , Liu Y , Huang Z , Li C , Wu D , Ruan R . (2022a). Multiple hydrolyses of rice straw by domesticated paddy soil microbes for methane production via liquid anaerobic digestion. Bioresource Technology, 354: 127184

[47] Wang L , Lei Z , Yang X , Zhang C , Liu C , Shimizu K , Zhang Z , Yuan T . (2022b). Fe₃O₄ enhanced efficiency of volatile fatty acids production in anaerobic fermentation of food waste at high loading. Bioresource Technology, 364: 128097

[48] Wang Y , Wang S . (2021). Soft sensor for VFA concentration in anaerobic digestion process for treating kitchen waste based on SSAE-KELM. IEEE Access: Practical Innovations, Open Solutions, 9: 36466–36474

[49] Xu R , Cao J , Wu Y , Wang S , Luo J , Chen X , Fang F . (2020). An integrated approach based on virtual data augmentation and deep neural networks modeling for VFA production prediction in anaerobic fermentation process. Water Research, 184: 116103

[50] Yang X , Du M , Lee D J , Wan C , Zheng L , Wan F . (2012). Improved volatile fatty acids production from proteins of sewage sludge with anthraquinone-2,6-disulfonate (AQDS) under anaerobic condition. Bioresource Technology, 103(1): 494–497

[51] Yang Z , Sun H , Kurbonova M , Zhou L , Arhin S G , Papadakis V G , Goula M , Liu G , Zhang Y , Wang W . (2022). Simultaneous supplementation of magnetite and polyurethane foam carrier can reach a Pareto-optimal point to alleviate ammonia inhibition during anaerobic digestion. Renewable Energy, 189: 104–116

[52] Yuan Q , Sparling R , Oleszkiewicz J A . (2011). VFA generation from waste activated sludge: effect of temperature and mixing. Chemosphere, 82(4): 603–607

[53] Zhang L , Mou A , Sun H , Zhang Y , Zhou Y , Liu Y . (2021). Calcium phosphate granules formation: key to high rate of mesophilic UASB treatment of toilet wastewater. Science of the Total Environment, 773: 144972

[54] Zhang Y , Li L , Ren Z , Yu Y , Li Y , Pan J , Lu Y , Feng L , Zhang W , Han Y . (2022). Plant-scale biogas production prediction based on multiple hybrid machine learning technique. Bioresource Technology, 363: 127899

[55] Zhao J , Wang D , Liu Y , Ngo H H , Guo W , Yang Q , Li X . (2018). Novel stepwise pH control strategy to improve short chain fatty acid production from sludge anaerobic fermentation. Bioresource Technology, 249: 431–438

[56] Zhong S , Zhang K , Bagheri M , Burken J G , Gu A , Li B , Ma X , Marrone B L , Ren Z J , Schrier J . . (2021). Machine learning: new ideas and tools in environmental science and engineering. Environmental Science & Technology, 55(19): 12741–12754

Conflict of Interest

The authors declare no competing interests.

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://doi.org/10.1007/s11783-023-1735-8 and is accessible for authorized users.

RIGHTS & PERMISSIONS

2023 Higher Education Press