Sensor-Based Optimized Control of the Full Load Instability in Large Hydraulic Turbines

Area of Science:

• Engineering
• Renewable Energy Systems
• Fluid Dynamics

Background:

• Large hydraulic turbines, like Francis turbines, are crucial for integrating renewable energy but face operational challenges.
• Off-design conditions can lead to dangerous unstable operating points, compromising grid stability and plant safety.
• Full load instability often limits the maximum output power of Francis turbines, restricting their operational range.

Purpose of the Study:

• To optimize sensor acquisition strategies for monitoring and controlling Francis turbines.
• To detect and prevent the unit from reaching critical full load instability points.
• To extend the operational range of Francis turbines by enabling safer operation closer to instability.

Main Methods:

• Field tests were conducted on a 444 MW Francis Turbine in Canada, acquiring over 80 signals from ten sensor types.
• A systematic approach was used to optimize sensor type, number, location, and acquisition frequency.
• Signal analysis focused on identifying indicators linearly correlated with oscillating power.

Main Results:

• Identified specific sensor-derived indicators linearly correlated with oscillating power.
• Determined an optimized acquisition strategy based on correlation characteristics, sensor simplicity, and required acquisition frequency.
• Validated that the optimized strategy allows for safe operation closer to the full load instability point.

Conclusions:

• An optimized sensor acquisition strategy is key to managing Francis turbine instability.
• The developed strategy enables extended operational range by allowing safe proximity to instability.
• An economical and easily implementable protection system based on this strategy is proposed for generic Francis turbines.