Minimizing Energy Consumption in Water Reuse: Digital+Physical Twin Approach

Minimizing Energy Consumption in Water Reuse: Digital+Physical Twin Approach P. Ghorbani*, **, B. Tarroja*, ***, N. Rezaei*, **, ****, A.M. Melin*****, S. Ghanem*****, K. Villez*****, and D. Rosso*, ** * Department of Civil and Environmental Engineering, University of California, Irvine, CA 92697, USA (E-mail: po[email protected]; [email protected]u; bta[email protected]; n[email protected]) ** Water-Energy Nexus Center, University of California, Irvine, CA 92697, USA *** Advanced Power and Energy Program, University of California, Irvine CA, 92697, USA (E-mail: [email protected]) **** AECOM, 999 W. Town and Country Rd, Orange, CA 92868, USA (E-mail: nad[email protected]) ***** Oak Ridge National Laboratory, Electrification and Energy Infrastructure Division, 1 Bethel Valley Road, TN 37920, USA (E-mail: [email protected]ov; melin[email protected]v; ghan[email protected]) Abstract High energy consumption is a major barrier to the efficient and environmentally sustainable management of water reuse facilities. This is exacerbated by fluctuating electricity prices and inefficient energy management. This study presents an integrated digital-physical twin framework to maximize the energy efficiency of water reuse processes. The digital twin models can forecast system dynamics and are used to infer optimal control actions, while the physical twin is used to evaluate the proposed control actions in real-world conditions. By leveraging time-of-use electricity pricing and dynamic process adjustments, the proposed framework achieved a 5% reduction in daily electricity costs without compromising system performance. Additionally, it enhances resilience by simulating and mitigating the impact of extreme events such as power disruptions. These findings demonstrate the potential of digital-physical twin integration in improving energy efficiency and sustainability in water reuse systems. Keywords Advanced water reuse; digital+physical twins; energy consumption optimization INTRODUCTION One critical challenge in water reuse facilities is the high energy consumption required for treatment and distribution (Beji & Lade, 2022). This high energy demand is exacerbated by the underutilization of dynamic electrical energy availability and fluctuating time-of-use (TOU) pricing, leading to inefficiencies in operational energy management. Optimizing process efficiency through advanced modeling, intelligent data-driven decision-making, and integrated energy management strategies is essential to address this issue. These approaches can enhance energy efficiency, reduce operational costs, and improve the sustainability of water reuse systems. To tackle this challenge, an integrated digital-physical twin framework is proposed. A digital twin, a highly detailed virtual representation of a real-world system capable of predicting outcomes with a high degree of accuracy (Ghorbani Bam et al, 2025), is employed to replicate process behavior and simulate anomalies caused by extreme events such as earthquakes, storms, and wildfires. Incorporating a resilience assessment model can provide a quantitative evaluation of system vulnerabilities, guiding resource allocation and system design to improve energy efficiency and operational resilience (Juan-García et al., 2021). Simultaneously, the physical twin (i.e., the pilot) serves as a controlled testing environment for validating the optimal operational setpoints derived from the digital twin. The distinguishing feature of this approach is the real-time connectivity between the digital and physical twins, enabling continuous adaptation and improvement. This framework is designed to maximize energy efficiency in water reuse processes while maintaining system performance by considering design parameters holistically rather than in isolation. A process-based model, combined with advanced data analytics, facilitates the exploration of various operating conditions, predicts outcomes in abnormal scenarios, and enhances the reliability of optimization strategies. Ultimately, this approach aims to create a more resilient and energy-efficient water reuse system. MATERIALS AND METHODS This study employs an integrated approach combining a digital twin, data-driven model training, and experimental design with a physical twin to optimize energy consumption in water reuse processes. SUMO Dynamizu was selected as the digital twin platform, where both the physical twin (pilot system) and the full-scale plant were modeled. To achieve energy optimization while maintaining system performance, two key digital twin-based strategies were implemented: • Accounting for Process Dynamics and Identifying Optimization Bottlenecks: The digital twin is essential for capturing complex process dynamics that constrain operational responses and for identifying specific processes or equipment that limit energy optimization. This allows for targeted improvements and more effective energy management. • Utilizing a Process-Based Model for Robust Optimization: Employing a process-based model allows exploration of out-of-normal operating conditions with minimal assumptions about equipment behavior. This approach ensures more accurate and reliable optimization strategies, even under diverse and challenging operational scenarios. To align with energy optimization goals, experiments were designed based on operational strategies incorporating TOU electricity tariffs. Various operational scenarios were tested, particularly analyzing the trade-offs between recirculation flow rate, recovery rate, and energy consumption. For digital twin training, pilot system experiments were conducted by deliberate and simultaneous variation of key operational parameters including feed flow, recovery ratio, and recirculation ratio. This information-focused experimental design ensures that the digital twin accurately reflects real system performance, enabling precise simulations and predictions under varying operational conditions. RESULTS AND DISCUSSION A 6-hour experiment was conducted with different permeate, concentrate, and recirculation flow rates to evaluate the trade-off between operational parameters and specific energy consumption (SEC). Figure 1 presents six distinct scenarios tested during the experiment, with hourly set points for permeate, concentrate, and recirculation flow rates. Figure 1. Time series of flow rates during the six-hour experiment (as the pilot system does not digitally record recirculation flow rates, the recirculation set points are presented in the figure). To compare the SEC cost in the advanced reverse osmosis (RO) system during the six-hour experiment with a baseline operation mode—defined by a recovery ratio of 75% and a recirculation ratio of 50%—the minimum energy costs were calculated using Southern California Edison’s TOU tariff structure. Figure 2 illustrates the normalized SEC cost comparison between the fixed operation and the adaptive adjustment of recovery and recirculation rates throughout the experiment. Figure 2. Normalized SEC cost over the 6-hour experiment (the pilot operates with a break tank after the ultrafiltration (UF), and drops in the first, second, and fourth phases correspond to UF feed pump shutdown due to the high volume in the break tank). Data collected from the physical twin during this experiment, along with additional experiments conducted under varying operational conditions, provided valuable insights into the trade-offs between operational parameters and SCE, which were instrumental in driving advanced energy and cost optimization strategies within the digital twin framework. Figure 3 demonstrates the diurnal cost optimization achieved through dynamic adjustments to the recovery ratio and recirculation flow rates based on Southern California Edison TOU tariff structures. Figure 3. Hourly cost optimization for a 24-window. The optimization framework applied here projects a 5% reduction in daily electricity costs while maintaining both the quantity and quality of water produced. As shown in Fig. 3, operating costs (and electricity consumption) fluctuate more hourly compared to baseline operations. The framework not only optimizes cost savings under normal conditions but also simulates the impact of adverse events, such as power disruptions from natural disasters, enabling proactive planning and mitigation. By operating the physical twin under stressed conditions, it generates data to predict and help avoid undesirable scenarios, enhancing the resilience of water reuse facilities. The full paper and the presentation include expanded results and discussion. CONCLUSIONS Integrating digital and physical twins offers a transformative solution for addressing energy challenges in water reuse facilities. Leveraging real-time connectivity and TOU tariffs, this approach enhances operational efficiency, as shown by the optimization framework’s 5% reduction in daily electricity costs while maintaining water quantity and quality. It also enables data collection for simulating and detecting the impact of adverse events like earthquakes, wildfires, and hurricanes. This capability supports proactive planning and mitigation strategies, ensuring operational continuity and minimizing disruptions. REFERENCES Beji, H., & Lade, M. (2022). Impact of Digital Transformation on Carbon Emissions Reductions in the Water Industry. In Lecture Notes in Energy (Vol. 86, pp. 117–127). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-86215-2 Ghorbani Bam, P., Rezaei, N., Roubanis, A., Austin, D., Austin, E., Tarroja, B., Takacs, I., Villez, K., & Rosso, D. (2025). Digital Twin Applications in the Water Sector: A Review. Water, 17(20), 2957. https://doi.org/10.3390/w17202957 Juan-García, P., Rieger, L., Darch, G., Schraa, O., and Corominas, L. (2021). A Framework for Model-Based Assessment of Resilience in Water Resource Recovery Facilities Against Power Outage. Water Research 202: 117459. https://doi.org/10.1016/j.watres.2021.117459