Project background

Next generation hydropower maintenance
The hydropower business is in competition with an increasing share of renewables in a changing energy market. One of the means to increase hydropower competitiveness is to ensure the production assets' reliability and to make operation and maintenance predictable and cost effective.

New opportunities through digitalisation
New concepts, methods and models - such as industry 4.0, machine learning, cyber-physical systems, internet of things and data mining - gives the hydropower business new opportunities. This will help with the transition from scheduled maintenance towards predictive and condition based maintenance. Some benefits are reduced number of manual inspections, continuous monitoring of technical condition and early warning of potential failures. Ultimately, this will reduce maintenance costs.

Digital transformation: New concepts, methods and models

Project focus
The MonitorX project focused on data analysis models and algorithms for condition monitoring and early warning of faults. One of the starting points of the project was the assumption that sensor and measurement data are not much used for planning and optimization of maintenance and refurbishment. Thus, a project with focus on model and algorithm development will be useful to motivate plant operators to accelerate and implement solutions for (online) data collection and analysis.

MonitorX mainly focuses on models based on machine learning and artificial intelligence. The models have been developed in a number of cases, where each case focuses on a specific type of hydropower component. Data from selected hydropower plants were used to develop the models. The models were tested with data from one or several power plants.

Project phases

Sketch of MonitorX phases and project timeline

The MonitorX project started with an initial phase where the status for collection of monitoring data and use of monitoring data in the power companies were analysed and where a list with relevant use cases were developed together with the industry partners. This included an overview of models for data analysis was performed. The MonitorX project phases and timeline is illustrated in the figure below. See MonitorX – Cases and Results for a case overview and a summary of MonitorX results.

Note that MonitorX did not focus on development of systems and platforms for data handling and analysis, such as solutions for data collection and storage. A number of different systems and platforms are available on the market, and it is assumed that these solutions serve most of the current needs, even though further developments may be required, e.g. within system security and interoperability (both towards data sources and towards systems and software for data analysis).