Out-of-step detection for synchronous generators using electrical power analysis and Durbin Watson testing

Area of Science:

• Electrical Engineering
• Power Systems Protection
• Computational Intelligence

Background:

• Shunt faults in Synchronous Generators (SGs) can cause significant electrical output fluctuations, leading to loss of synchronization with the power network.
• Diagnosing power quality instability and distinguishing between synchronous and asynchronous generator operation is crucial for grid stability.
• Existing protection strategies may not adequately anticipate or rapidly detect Out-of-Step (OOS) conditions following faults.

Purpose of the Study:

• To develop and validate an intelligent relaying strategy for anticipating and detecting generator Out-of-Step (OOS) situations.
• To identify sudden variations in key electrical parameters during OOS events for accurate fault diagnosis.
• To ensure rapid tripping of protective relays to prevent severe grid disturbances.

Main Methods:

• Utilized computational techniques as a foundation for an intelligent relay to detect OOS events.
• Developed a protection strategy that monitors phase voltage, current, active power, reactive power, and power angle for OOS signatures.
• Validated the method using a power system model with real component data in ATP, with algorithm analysis performed in MATLAB.

Main Results:

• The protection strategy successfully identified OOS events, triggering protective relays to trip generator circuit breakers.
• The system remained inactive under stable synchronization conditions, demonstrating selectivity.
• OOS conditions were announced rapidly, preceding the second pole-slipping occurrence, and the algorithm proved robust during stable power swings.

Conclusions:

• The proposed computational technique provides an effective and robust method for detecting generator Out-of-Step conditions.
• The strategy offers protection redundancy and accurately estimates instability time and frequency rate.
• This intelligent relaying approach enhances power system reliability by enabling swift and precise detection of generator instability.