Friday, August 8, 2014 at 2:00 P.M.
“Lifetime Estimation of IGBTs in a Grid-connected STATCOM”
Reliability is a matter of concern in power systems integrated with Smart Grid (SG) technology, and there is a need to study the individual reliability of grid-connected devices to predict their impact on the grid, also emphasized by North American Electric Reliability Commission (NERC). The overall reliability is based on the reliability of the converter systems for Flexible Alternating Current Transmission Systems (FACTS), and STATCOM, STATic COMpensator, in particular. According to MILitary-standards (MIL-std 217), the Insulated Gate Bipolar Transistors (IGBTs) are the second most unreliable devices causing failure of the inverter, after capacitors, in a power converter.
This research focuses on the reliability of power semiconductor devices, and IGBT devices in particular, used in the power electronics integrated with power systems. Since most of the common failures (wire bond and solder fatigue) are caused by thermo-mechanical stresses, the methodology of lifetime estimation starts with temperature estimation, cycle counting based on rainflow algorithm, and finally degradation calculation based on linear accumulation model is performed.
An optimum number RC based thermal model, including the influence of encapsulant is proposed for temperature estimation of IGBTs and diodes in power modules. Temperatures of IGBT are estimated and presented when the STATCOM is operating in power factor correction mode. A novel rainflow algorithm with faster execution time is proposed for cycle counting. An operation parameter based lifetime model for a wirebond is proposed to estimate lifetime from Finite Element Analysis (FEA). Finally, the lifetime of the IGBT is estimated during STATCOM operation using real-time load profiles for power factor variation.
Knoxville, Tennessee 37996 | 865-974-1000
The flagship campus of the University of Tennessee System