Single-turn sensitive stator inter-turn fault detection of induction machines
Catastrophic failure of the electric machines can result from stator inter-turn faults even at their very incipient stage i.e.. single-turn fault. Consequently, fire and explosion, loss of human life and property, extended downtime of the equipment, increased cost of repair and heavy financial losse...
| Main Author: | |
|---|---|
| Other Authors: | |
| Language: | English en |
| Published: |
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1828/2499 |
| Summary: | Catastrophic failure of the electric machines can result from stator inter-turn faults even at their very incipient stage i.e.. single-turn fault. Consequently, fire and explosion, loss of human life and property, extended downtime of the equipment, increased cost of repair and heavy financial losses in the industries may take place. As a recent trend, online fault diagnosis of the electric machines that are employed in critical applications has been considered of paramount importance since frequent outage of the machines for the purpose of testing cannot be recommended. In this thesis, a very accurate diagnostic scheme has been developed to unambiguously detect single-turn faults on line in squirrel cage induction machines by detecting positive and negative sequence line current third harmonic components.
Initially, inadequacy has been identified in a diagnostic scheme based on negative sequence quantities of the machine and critical improvements have been realized to
suppress the effects of changing supply unbalance. However the modified method fails to detect faults involving one turn short. The feasibility of computing positive and negative sequence line current third harmonic components have been proved by conducting space vector analysis. Subsequently, a detailed description upon the introduction of these line current third harmonic components is given that accounts for the effects of air gap permeance harmonics. MMF harmonics under supply unbalance, internal asymmetries as well as fault conditions. Least-square method is employed to discriminate fault signatures from those arise from other abnormal operating conditions. The implementation of the scheme is carried out on both simulated and experimental machines. Mathematical models of the induction machines have been developed which took different induction machine operating conditions into consideration. The machine has been simulated to verify the differences among different complex coefficients by computing complex line current spectra fault signatures with ideal (harmonics-free) three phase voltages input as well as experimentally collected voltages input. Further in an attempt to test the effectiveness of the diagnostic scheme under practical conditions, stator inter-turn faults associated with varying fault severities have been introduced to the line and inverter fed machines under different stator configuration. It has been demonstrated from experimental results that proposed fault signatures can achieve unambiguous single-turn fault detection. In addition the need of different coefficients for computing the positive and the negative third harmonic related fault signatures have also been demonstrated. |
|---|