Dodola - Predictive control for efficient and flexible water distribution system operation
The project aims to foster stronger collaboration between research institutions (University of Zagreb Faculty of Electrical Engineering and Computing) and companies (KONČAR - Digital d.o.o. and Vodne usluge d.o.o.). By enhancing their capacities for research, development, and innovation, the project will conduct applied research on predictive control systems designed for the efficient and adaptable operation of water distribution networks.
The key objectives are to tackle critical challenges in the water distribution sector, including reducing water losses and minimizing electricity costs, by applying advanced methodologies that outperform current approaches.
The target groups include the project participants (the applicant and partners), external research institutions and companies, international researchers, and potential customers of the Dodola solution.
Contact person: Tamara Hadjina (tamara.hadjina@koncar.hr)
Project description
Dodola is a collaborative applied research project led by KONČAR - Digital d.o.o., Vodne usluge d.o.o., and the University of Zagreb Faculty of Electrical Engineering and Computing. The project encompasses both fundamental and industrial research, focusing on technology readiness levels (TRL) 1-4 to develop the innovative Dodola product. This solution is centered around predictive control systems, aimed at optimizing the efficiency and adaptability of water distribution networks.
Dodola offers an innovative approach to minimizing water losses and optimizing electricity usage in water distribution systems.
Traditional water distribution systems are often operated based on manual flow and pressure settings for individual pumps or valves, typically set by the operator’s experience or predefined schedules. Autonomous control loops are often added to regulate pump operations, ensuring water levels in reservoirs—and consequently, system pressures—remain within specified limits. Most systems are monitored through basic SCADA (Supervisory Control and Data Acquisition) systems, enabling operators to observe key parameters and make ad-hoc interventions when needed.
This type of system operation is quite conservative, relying on predefined control actions to ensure adequate pressure levels at system endpoints. However, it overlooks operational costs associated with energy consumption and water losses and fails to leverage the potential for participation in power grid ancillary service markets. Such opportunities arise from the flexibility inherent in managing water reservoir levels.
Water distribution systems are inherently dynamic, requiring consistent pressure regulation across the network while minimizing energy consumption and water losses. These factors make them prime candidates for predictive control and optimization techniques. The aim of these techniques is to determine the optimal settings for pumps and valves over a defined period. A phased implementation approach can be adopted by initially providing operators with optimized recommendations, supported by key performance indicators that validate the proposed operational strategies.
Implementing a computer-based decision-making system necessitates the creation of a comprehensive mathematical model of the water distribution network. This model serves as the foundation for optimization algorithms; however, many water supply companies lack the necessary expertise in modelling, optimization, and programming.
Furthermore, each water distribution system is unique, making the development of a detailed and accurate model particularly challenging. The model must also account for loss calculations, which require extensive data monitoring across multiple points in the network.
As part of this project, the design and conceptual development of an autonomous predictive control system will be carried out. The system's performance will be tested using a lab-scale simulation platform integrated with an existing SCADA system, which is designed for easy upgrading and integration into water distribution automation systems.
Another advantage of predictive control in water systems is the potential to leverage the inherent flexibility of water storage facilities to enable demand response, thereby further reducing electricity costs. These savings stem partly from pump operation in alignment with electricity tariffs and partly from financial incentives for flexible energy consumption.
Such Flexible water distribution systems also facilitate greater integration of renewable energy sources, an important indirect benefit of the Dodola solution. Given the intermittent nature of renewable energy production, the responsibility of balancing supply and demand is increasingly placed on consumers through demand response mechanisms. Encouraging consumers to adjust their energy consumption based on renewable energy availability can significantly enhance the integration of these sources into the grid.