Health Monitoring Applications for Small Wind Turbines

Open Access
Author:
Forshey, Brenton Ross
Graduate Program:
Aerospace Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
July 29, 2001
Committee Members:
  • Dennis K Mc Laughlin, Thesis Advisor
  • Dennis K Mclaughlin, Thesis Advisor
  • Karl Martin Reichard, Thesis Advisor
  • Susan W Stewart, Thesis Advisor
Keywords:
  • bearing
  • detection
  • fault
  • condition
  • monitoring
  • health
  • energy
  • wind
  • imbalance
  • structural health monitoring
Abstract:
Costly failures have plagued industrial-scale wind turbines in recent years, and these failures have spawned the development of systems designed to provide continuous data on the health of these machines. Though they vary in implementation, the common goal of these systems is to detect potential faults before they lead to catastrophic failures and extended downtime of the turbine. The most basic condition monitoring systems typically employ vibration sensors placed at key locations in the machine that can detect faults that develop within the main bearings. Torsion and bending sensors on the tower are also used to identify and distinguish various types of blade imbalances, be they aerodynamic or mechanical in nature. The purpose of this investigation was to develop methods for fault detection on a residential-scale wind turbine. Two separate fault detection experiments are presented in this thesis. The purpose of the first experiment was to detect and qualify the severity of an inertial imbalance seeded on a Whisper 500 turbine in the field. Such imbalances are typically the result of ice buildup during normal operation. Through the observation of passive vibration spectra from a tower-mounted accelerometer, two separate imbalance cases were able to be successfully detected and differentiated from the baseline case. A second test was conducted on a dynamometer with the purpose of detecting a seeded fault within one of the wind turbine’s main bearings. It was shown through the use of envelope analysis that lab-grade accelerometers could successfully detect vibration frequencies that indicated the presence of a seeded outer race fault in the Whisper 500’s rearmost bearing.