2TB of data processed each month

Up to 1% Increase in availability

Detecting failures Before they happen



Hydropower, as one of the largest, most flexible and most efficient renewable energy sources available, contributes to grid stability thanks to its scale of production and its flexibility. Its flexibility also makes it a key technology for integrating other intermittent renewable sources of energy into the grid. In this context, Hydro utility companies are changing the way they operate their plants, switching from baseload to more flexible power production. Intelligent condition monitoring and diagnostics become crucial.

To adapt to this changing production mode, EDSB (Energie Développement Services du Briançonnais) is looking at expanding the time between plant overhauls (TBO) and shortening the mean time for repair (TTM).

A monitoring system intelligent enough to track the health of the plant in order to detect failures before they happen- so that the plants can be repaired at the best possible time to minimize downtime- is of very high value. Up until now, there had been no such system on the market.


In December 2015, Pont Baldy - a hydropower plant operated by EDSB in South East of France - was connected to a new generation Condition Monitoring System (CMS) called iCMS.

It remotely collects real-time data, that GE analyses to improve diagnostics and prognostics on faults in the plant that the monitoring system has identified.

The iCMS has been collecting and analyzing almost 2 Terabytes of raw data per month that are combined with 3 years worth of previously collected data. Thanks to their analysis turned into predictive models, faults and maintenance operations can then be identified automatically through a Human-to-Machine Interface (HMI) that gives recommendations on what to do for each diagnosis. Faulty components are easily detected and the HMI helpfully brings up any related documents, manuals and reports to make the repairs process as smooth and as fast as possible.

The Pont Baldy plant had for example a bearing that was running hot and iCMS diagnosed the cause and analyzed that no action was needed, saving time & expense of unnecessary repair.

Overall, the iCMS can bring up to 1% extra availability for the power plant.

GE’s new iCMS, the ‘on premise’ part of the GE Asset Performance Management (APM) solution, uses machine learning and innovation to turn monitoring and maintenance into a successful story.