« Retourner aux résultats de recherche

• Citer

Water-Energy Nexus Optimization:: Technical and Experimental Validation of Energy Recovery and Flexibility Services in Water Supply and Distribution Networks. A Case Study on the Brussels-Capital Region

par Nuc, François
Président du jury Parente, Alessandro
Promoteur Hendrick, Patrick
Publication Non publié, 2026-06-05

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Water-Energy Nexus Optimization: Technical and experimental validation of energy recovery and flexibility services in water supply and distribution networks (WSDN). A Case Study on the Brussels-Capital Region.This doctoral thesis investigates energy optimization opportunities within the Water Supply and Distribution Network (WSDN) serving the Brussels-Capital Region (BCR), managed by the company Vivaqua. This research addresses the critical challenge of rising electricity costs that European countries are currently facing, which affects all industries and households, including water network operators.Acknowledgement: This research was conducted within the framework of the FlexWatter project, a Belgian federal research project dedicated to the optimization of urban water distribution systems. The project brings together academic and industrial partners, including Université libre de Bruxelles (ULB), University of Liège (ULiège), Ghent University (UGent), Vivaqua, Pepite, and ICEED. The author gratefully acknowledges the support, collaboration, and data sharing provided by all partners.Structured in three complementary yet independently readable parts, the research combines a comprehensive literature review, a large-scale industrial data analysis with energy optimization strategies, and an experimental validation on a dedicated Hydropower test bench.Part I establishes theoretical foundations through extensive review of WSDN design, operation, and optimization methodologies. Key concepts including network topology (branched vs. looped configurations), sectorization strategies (DMAs and PMAs), hydraulic modeling, multi-objective optimization algorithms, and real-time control systems are systematically presented. This section underlines a key fact: WSN and WDN are not the same infrastructures and they do not have the same behavior. In scientific literature WSDN are almost exclusively referred as WDN. This literature review shows, among other facts, that WSN and WDN have neither the same purposes nor the same structures. This pedagogical approach bridges knowledge gaps between water engineering and energy systems disciplines.Part II presents a detailed analysis of the Brussels-Capital Region WSDN structure, encompassing four production centers (Tailfer, Modave, Spontin, Mons-Havré), extensive supply infrastructure (742 km), and urban distribution networks (2,370 km) serving Brussels. Analysis of four years of operational data (2019-2022, representing 10.7 billion sensor measurements at 1 Hz) revealed a critical paradox: despite 30% reductions in both water production and electricity consumption, operational costs increased by 28% due to European electricity market dynamics. This paradox underscores the urgent operational necessity for the BCR WSDN utility to explore multiple strategies to reduce electricity-related costs.Four energy optimization strategies were evaluated:1. Electrical flexibility via pumped-storage hydropower (PSH): A 1.8 MW turbine installation between Tailfer and Bois-de-Villers reservoirs could generate 1,330 MWh annually, but operational constraints (daily water management cycles conflicting with TSO flexibility activation requirements) and extended ROI periods (8+ years under favorable market conditions) limit viability.2. Load shifting optimization: AI-driven self-learning algorithms (Pepite Data Maestro platform) demonstrated €420,000 cost savings over four years through intelligent pump scheduling aligned with Day-Ahead electricity market price signals. This strategy requires zero CAPEX, operates independently of market price volatility, and proved immediately implementable.3. Energy recovery at pressure-regulating valves: Systematic analysis identified viable installation locations at Mazy WSN station (valves V2, V7), where cross-flow turbines or pumps-as-turbines (PaT) could recover 2,531 MWh annually . Critical findings revealed that network behavioral characteristics, particularly square-wave pressure variations from Callois reservoir topology and cascading effects throughout WSN/WDN, directly influence technology selection (PaT preferred for pressure variations; cross-flow turbines for flow variations).4. Battery Energy Storage System (BESS): This section explores the potential of harnessing energy dissipated at Mazy through a turbine to power a Battery Energy Storage System (BESS). Three scenarios with different configurations are evaluated to structure viable business plans over a 20-year horizon. All analyses are based on actual 2025 ELIA day-ahead market data. Part III describes design, construction, calibration, and experimental validation of a dedicated Hydropower test bench at University of Liège, replicating a WSDN structure. The installation integrated industrial components: JLA cross-flow turbine (1.8 kW), Cla-Val pressure-regulating valve (DN200), and KSB pump-as-turbine (1.5 kW), with comprehensive instrumentation (torque meters, flow/pressure/level sensors) and LabView-based data acquisition (1 Hz sampling).Three experimental campaigns validated technical feasibility:• Experiment 1 replicated WSN square-wave pressure variations, confirming PaT adaptation to backpressure fluctuations (1,500W → 1,000W → 1,500W cycles, maintaining 1,200W average).• Experiment 2 demonstrated cross-flow turbine exceptional flow-tracking capability (3-second stabilization) and variable-speed operation benefits (10% efficiency improvement, eliminating power consumption during low-flow conditions).• Experiment 3 revealed complex hydraulic interactions in parallel turbine-valve configurations, identifying flow distribution dynamics and pressure coupling effects critical for real-network implementation.Principal conclusions: Water and energy are intrinsically linked, making the Water–Energy Nexus a central component of modern infrastructure. In a context of elevated day-ahead electricity prices in Europe, energy optimization in WSDNs has become essential. Several strategies were assessed for the Brussels-Capital Region. Load shifting appears as the most effective short-term solution, requiring no investment and offering robust, market-independent returns with limited network impact. Energy recovery at selected valves represents a relevant secondary option. In contrast, pumped storage hydropower shows operational incompatibilities with water network constraints, while BESS remains conditionally viable due to high capital costs and market exposure.Long-term data analysis (3–4 years) proved critical to capture network variability and ensure robust decision-making. The experimental test bench validates key aspects of energy recovery strategies and confirms the relevance of Pump-as-Turbine (PaT) systems for pressure variations, and cross-flow turbines for flow variability.Overall, this work contributes both to academic research and industrial practice by providing a data-driven and experimentally validated framework for WSDN optimization The methodology developed in this research is not limited to Brussels-Capital Region, and is highly replicable and scalable to any WSDN infrastructure globally.