Challenge
Solar plants are central to a low-carbon future for our societies. However, an energy system reliant on renewable energies requires sustained and predictable performance, a constant challenge for solar installation operators facing curtailment from the central grid.
Innovative Solution
In order to address the issue of curtailment and lost energy produced by solar installations in Vietnam, Bouygues Energies & Services’ Solar Energy team developed an artificial intelligence algorithm capable of predicting plant performance. The AI learns from a variety of datasets, ranging from weather patterns to solar plant configuration, and predicts with over 90% certainty the performance expected from the installation at any given point.
Clients can then use these data points in order to explore solutions for the use of solar power lost through curtailment.
Benefits
- Accurate prediction of solar plant performance on a day-to-day basis.
- Dashboards for monitoring the performance of the solar plants
- Automatic calculation of deviations between model and measured values to define underperformance
Key Figures
- This solution was tested in Vietnam on a power plant with a capacity of 250 MWp.
Maturity Level
Level 1: Proof of Concept oder lab test
Level 2: Tested under real life conditions
Level 3: Commercial solution