Hydroelectric powerplants (HEPs) are the backbone of the electricity grid in Norway, contributing to more than 90% of the total power production. Owing to the exceptional Norwegian terrain and water resources, the HEPs have significantly contributed to Norway's industrial development over the past century. As EU gears towards increased reliance on renewable energy, with 32% of total energy requirement from renewable energy sources by 2030, Norway has the capacity to play a greater role in the EU grid as the ‘green battery of Europe. The electricity grid of the future needs reliable energy sources that can rapidly meet the capacity when for example, the wind power is low, and store excess energy produced when wind speeds are high. Norway’s abundant hydropower has the capacity to meet these requirements as power production can be increased rapidly and energy can be stored e.g. by pumping water back to reservoirs. To achieve this, the HEPs should have high reliability and efficiency to ensure a stable grid. Continuous monitoring of the condition of the HEP critical equipment like turbines and generators, and efficiency ensures that the HEPs are operated profitably and without unwanted shutdowns.
The project PHMHydro aims at building the fundament for such performance and health monitoring architecture for HEPs that is scalable to multiple HEP components and HEPs. This includes evaluating novel sensing methods, developing new algorithms using physics and artificial intelligence (AI), and an architecture for handling these algorithms in cloud/fog deployment. The project is collaborative research from the University of Agder (UiA), NORCE Norwegian Research Centre AS, and the industry partners Agder Energi AS and Volue Industrial IoT AS. The developed solutions will be evaluated in a pilot HEP with Agder Energi AS, and Volue Industrial IoT AS will be involved in instrumentation and evaluation of the sensing technologies towards their suitability for Hydropower plants.
The project PHMHydro is focused on improving the longevity and profitability of Hydro Electric Powerplants (HEPs) by accelerating the penetration of data analytics in operations and maintenance (O&M). The project aims at closing the knowledge gaps related to:
- Insufficient condition monitoring (CM) and predominantly conventional maintenance planning
- Ineffective use of SCADA data
- Lack of systematic fatigue assessment in powerplants
- Lack of plant-wide data analytics
- Lack of performance assessment and reliability-based forecasting.
The project will reach its goal through a three-pronged approach:
1) Develop a generalised architecture that enables aggregation of multi-modal data from CM, SCADA and other sources and modularized Health Assessment (HA) and Performance Analytics (PA), suitable for deployment across multiple HEPs.
2) Develop HEP-specific multi-sensorial CM solutions for the two most critical systems, hydro turbine and generator, using cost-effective retrofit-friendly sensing solutions, and leveraging the power of artificial intelligence (AI), machine learning (ML) and deep learning (DL), fuse CM and SCADA data for HA.
3) Develop continuous online PA forecasting safe production capabilities given health degradation for effective management of HEP operations and minimizing supply risk in contractual guarantees.
Through PHMHydro, the modular solutions developed in 2) and 3) will enable demonstration of the data analytics platform 1), and the collaboration with HEP owner (Agder Energi) as well as instrumentation supplier (Volue Industrial IoT AS) will ensure onsite testing and validation, ensuring high technology readiness level (TRLs). The advantage of such a generalized architecture developed in PHMHydro goes beyond the activities of this project as it fosters rapid expansion of CM to other systems as well.