A Stage-Wise Literature Review Toward Smart Digital Twin Developments of Wastewater Treatment Plants

TEchMA 2025 New Frontiers in MechanicalEngineering recuperarportugal.gov.pt A Stage-Wise Literature Review Toward Smart Digital Twin Developments of Wastewater Treatment Plants Abstract WWTPs are essential for safeguarding public and environmental health. However, they are energy-intensive infrastructures, consuming over 1–3% of global electricity and more than 233 GWh annually in Europe [1]. These challenges are compounded by increased inflow variability due to climate change and more restrictive environmental regulations. To address these issues, data-driven modelling and smart digital tools have become pivotal in enhancing WWTP operational efficiency and sustainability. This systematic review explores recent advances in ML, the IoT, and Digital Twin technologies applied to WWTPs. Recent studies have revealed two primary modelling categories: (1) system simulation and (2) energy optimisation. Hybrid models that combine physics-based simulations with ML techniques, such as neural networks, reinforcement learning, and AI-CFD, show improved prediction accuracy and generalisability. While these technological advances show promising results, the practical, widespread implementation of fullscale digital twins encounters significant obstacles. This review identifies, amongst others, a significant research gap: the lack of integrated, cross-stage optimisation strategies that bridge localised model insights and global plant performance. The novelty of this work lies in its stage-wise analysis of the literature, providing a structured perspective that helps identify specific WWTP stages where optimisation is most effective. It also serves as a foundational assessment for the design of Smart Predictive Digital Twins (SPDTs) that combine real-time data, ML-driven modelling, and modular architectures to enable adaptive, efficient, and resilient WWTP operations. Introduction Wastewater treatment plants (WWTPs) are among the most energy-intensive components of urban infrastructure, with aeration alone accounting for over 50% of total energy consumption [2]. In recent years, the increasing frequency of extreme rainfall events, driven by climate change, has led to frequent sewer overflows. This not only raises the energy demands and operational costs for treating excess inflow but also put at risk compliance with environmental regulations, potentially resulting in untreated discharges and legal penalties. Fig 4 / Most used ML architectures for individual stage simulation of a WWTP. Fig 2 / Discretization of a WWTP different stages and associated number of related works. Each numbered box represents the number of research studies identified in the literature that apply datadriven techniques, at that specific treatment stage. Fig 3 / Number of wors by research scope on the field of WWTP modeling. Sara Mota, António Andrade-Campos Acknowledgement This research was supported by the FEDER and Regional Operational Program of the Center Region (CENTRO2030) within project I-ReTiS-LeaksD&Op nº 17304 (CENTRO2030-FEDER-01177300) and through the Portuguese Foundation for Science and Technology (FCT), supported by the Recovery and Resilience Plan (PRR), within project I-ReTiS-Leaks (2024.07270.IACDC). This work is funded by national funds through FCT – Fundação para a Ciência e a Tecnologia, I.P., under the project/support UID/00481 – Centre for Mechanical Technology and Automation (TEMA). References [1]I. C. for Resource Recovery from Water, “Circular economy: Tapping the power of wastewater.” [2]G. Sabia et al., “Energy saving in wastewater treatment plants: A methodology based on common key performance indicators for the evaluation of plant energy performance, classification and benchmarking”, Energy Conversion and Management, vol. 220, 2020 [3]J. F. de Canete et al., “Control and soft sensing strategies for a wastewater treatment plant using a neuro-genetic approach”, Computers & Chemical Engineering, vol. 144, 2021. [4]ERSAR - Dados de base. Retrieved March 16, 2024, from https://www.ersar.pt/pt/setor/factos-e-numeros/dados-de-base [5] Duarte, M. S., Martins, G., Oliveira, P., Fernandes, B., Ferreira, E. C., Alves, M. M., Lopes, F., Pereira, M. A., & Novais, P. (2024). A Review of Computational Modeling in Wastewater Treatment Processes. Doi: https://doi.org/10.1021/acsestwater.3c00117 Fig 1 / Challenges and associated numbers for wastewater treatment plants, in Europe, for the year of 2022. [4] How? Regulatory Compliance Climate change Storm water overflows Virtual Entity Physical Entity How can data-driven techniques help to improve decision-making and energy use in wastewater management? 233 GWh of energy consumed for treatment yearly 48.57 Million € of savings associated with undue discharges 110% Additional water volume treated Individual process simulation 56% 24% 12% 8% Architecture Neural NetworkBased Random ForestBased SVM-Based Regression-Based Best ForKey FeaturesSoftware Comprehensive plant modelingAdvanced dynamic simulation, user-friendly interface GPS-X Industrial & municipal WWTPs Robust activated sludge modeling, integrated processes BioWin Process optimizationFlexible dynamic simulation, control strategy tools SIMBA# Educational & small-scale plantsFreeware, sludge & recycling modelingSTOAT Control strategy developmentDetailed process design, advanced optimization tools WEST Table 1 / Overview of software solutions for WWTP simulation. Methodology This work investigates how data-driven techniques, including Machine Learning, Digital Twins, and IoT, can enhance decision-making and energy efficiency in WWTPs. By conducting a systematic, stage-wise review of recent studies, the research identifies how these technologies support predictive control, optimize energy-intensive processes like aeration, and enable real-time process adjustments. A commercial review was also conducted in order to identify some practical implementation A preliminary treatment, removes coarse debris and grit to protect downstream equipment. In the primary treatment stage, heavier solids settle by gravity in primary sedimentation tanks. The core of the treatment occurs during secondary treatment, where biological processes, primarily using microorganisms, degrade organic pollutants, followed by secondary sedimentation to separate biomass from treated water. Tertiary treatment then further refines the effluent through advanced filtration and disinfection methods (e.g., UV) to meet high-quality discharge or reuse standards. Finally, effluent monitoring ensures that water released into the environment complies with all regulatory requirements. Results Aeration and sludge removal Primary sedimentation Bar screens and grit removal Biological treatment Global model of WWTP Local model 6 5 4 2 1 Secondary sedimentation UV Water disinfection 10 Effluent flow Influent flow Majority of studies focus on the entire WWTP rather than isolated processes. 0 1 2 3 4 5 6 7 8 Energy optimization Effluent quality parameters No. works per scope Entire WWTP Individual process Conclusion Stage-wise analysis reveals that smart technologies can enhance energy efficiency, particularly in aeration, as it improves effluent quality, and enables predictive maintenance. Real-world adoption remains limited due to challenges such as poor data quality, limited expertise, and high computational demands. The development of Smart Predictive Digital Twins (SPDTs) with modular, hybrid architectures are proposed as a promising path toward more adaptive, efficient, and sustainable WWTP operations. Analitical Hybrid Data driven •Review from Duarte et al. [5] •Physics informed ML •Digital twins •ANN •RNN •CNN Simulation Scope Energy Optimisation Operational forecasting Individual Process Entire Facility •Full plant condition •KPIs •MPC •Inflow prediction •Emissions •Aeration •Pumps •Mixing •Nutrient recovery •Preliminary treatment •Primary treatment •Secondary treatment •Tertiary treatment •Discharge WWTP stages Fig 4 / Structured summary of WWTP simulation literature: stages, research scope, and applied techniques The highest research concentration is seen in secondary treatment, particularly the aeration process, due to its high energy demand.