EEE/Solid state transformers could be the next big thing in power electronics

Author Topic: EEE/Solid state transformers could be the next big thing in power electronics  (Read 394 times)

Offline didarul alam

  • Full Member
  • ***
  • Posts: 109
  • Test
    • View Profile
There's been a lot of talk in the past few years about coming up with a solid-state version of the distribution transformer that now sits on utility poles in neighborhoods throughout the land. A solid-state transformer (SST) would be at least as efficient as a conventional version but would provide other benefits as well, particularly as renewable power sources become more widely used. Among its more notable strong points are on-demand reactive power support for the grid, better power quality, current limiting, management of distributed storage devices and a dc bus.

It is starting to look more likely that we'll see a practical SST design as GaN and SiC power transistors with higher current and voltage ratings start coming to market and their prices drop. But a practical SST design could have an impact extending well beyond transformers for electrical utilities.

One example of where SSTs could also find use is in variable-frequency drives for big induction motors. In that regard, Siemens Industry Inc.'s Drive Technologies Div. in New Kensington, Pa. is keeping an eye on SST work now underway at North Carolina State University's FREEDM System Center for smart grid research. Siemens Principal Engineer Mark Harshman says use of SSTs in the medium-voltage motor drives that Siemens makes could conceivably reduce the size of the VFDs by 30% and have similar beneficial effects on their overall efficiency levels.

There have been several topologies suggested for SSTs but most being evaluated today are based around the idea of a dual active bridge (DAB) converter. A DAB uses a power bridge to modulate the incoming ac waveform into a high-frequency square wave. The square wave gets passed through a small high-frequency transformer to another power stage. This converter demodulates the square wave and sends it to another inverter which produces low-voltage ac.

This scheme still uses a conventional transformer, but one optimized for higher frequencies (typically about 1 kHz). This makes it much smaller and lighter than transformers optimized for ac line frequencies.

The high-frequency transformer gives the SST galvanic isolation. It also has some leakage inductance in its primary and secondary windings, which also helps synthesize soft switching. During switching transients, transformer current resonates with the capacitors in parallel with switching devices, limiting the dv/dt and di/dt across the switches, thus reducing switching loss and boosting power efficiency.

The fact that DAB converters have a symmetrical circuit configuration lets them handle bi-directional power flows, important when it comes to renewable sources sending power back up the grid. The power flow of a DAB converter can be controlled by varying the phase shift between those two bridges which changes the voltage across the transformer leakage inductance. Power transfers from the leading bridge to the lagging bridge.

One of the difficulties in fabricating a SST is that the 7.2-kV line voltages that characterize distribution power lines exceed the operating voltage of today's IGBTs, 6.5 kV. So multiple devices must be used in series to keep below the operating maximum. The NC State prototype, for example, uses a topology that includes a seven-level cascaded H-bridge for the high voltage rectifier stage.

There are other difficulties as well. One is that the minimum current rating for the 6.5-kV IGBTs is 200 A. This is too large for the 20 kVA transformer NC State is building because the input current is only about 3 Arms. Thermal issues also affect the SST's operation, which has forced NC State researchers to come up with special packaging for their 25-A IGBTs. Additionally, the team had to come up with a way to isolate IGBT drivers for both power supply and gating signals.

To sense the 7.2 kVac voltage, the researchers devised a sensor that was compact and which incorporated high-voltage isolation because existing models were too large and not isolated from the high input voltage. Finally, they had to get around the fact that the insulation capability for 6.5-kV IGBT is 10.2 kV, but the high-voltage-side dc bus voltage is 11.4 kV. They ended up floating the heatsink for each 6.5-kV IGBT while maintaining ample clearance and creepage distance between the heatsinks. To keep the voltage across input inductor down to manageable levels, the team built eight identical inductors and put them in series so the maximum voltage stress for each of them is just 0.9 kV.

Researchers have also developed a prototype using 15 kV SiC MOSFET/JBS diodes. They are not trying to identify other major issues related to implementing a high-voltage system using SiC power devices, including the challenges in designing a system to support high dV/dt and dI/dt, and to design an efficient and compact high-frequency transformer.

More info from the FREEDM project:

Md. Didarul Alam

Offline sutapa.eee

  • Full Member
  • ***
  • Posts: 205
  • Test
    • View Profile
Good read.

Offline saikat07

  • Hero Member
  • *****
  • Posts: 630
  • Test
    • View Profile
    • My Web Address
Thanks for sharing
Senior Lecturer,
Department Of Electrical and Electronic Engineering
Faculty of Engineering,
Daffodil International University.