Are We at the Start of an Electrical Insulation Revolution: Nanodielectrics
There has been no radical change in the insulation materials used in motor and genera‐tor windings since the 1960s. Epoxy‐mica, aramid papers
(NomexTM), epoxy glass laminates and polyamideimides, the main insulating materials found in today’s rotor and stator windings, have in fact been in general commercial use for over 50 years. Of course, there have been some improvements in the properties of these materials over time, and even some creepage of material thermal ratings, but the basic materials remain the same. What advances there have been have largely been due to manufacturing methods – such as the large scale adoption of global VPI and the widespread application of robotics for applying tapes and processing [1].
Revolutionary change may be on the way with the commercial introduction of “nanodielectrics”. The recent issue of the IEEE Electrical Insulation Magazine was devoted to nanodielectrics [2]. This is a term, first coined by Dr. Michel Frechette of Hydro Quebec, for polymers that have nano particles dispersed within the polymer. “Nano” refers to the size of these particles, their largest dimension is in the order of 10s or perhaps 100s of nanometers (nm). Nanodielectrics are insulation materials where the nano‐sized particles have been evenly dispersed throughout the polymer. The nano particles to date have usually been silicon dioxide (sand), alumina or similar oxides. The promise of nanodielectrics is that they can produce a 10 to 100 times improvement in voltage endurance – that is, they seem to be exceptionally resistant to partial discharge (PD). In applications such as power cable and transformers, this means the insulation can be made much thinner for the same lifetime. For high voltage stator windings, the development of effective dielectrics may permit the elimination of mica from the ground and turn insulation, which may greatly reduce the complexity of manufacturing bars and coils. Nanodielectrics also tend to have much lower thermal impedance ‐ which means it is easier to conduct the I2R heat to the stator core, lowering the conductor temperature (or less copper can be used for the same temperature).
Research on nanodielectrics has been ongoing for over 20 years – with few commercial products to show for it. In the first decade, research was concentrated on how nanodielectrics worked, since it was not initially clear why such a small amount of very tiny particles creates such a big effect (usually the nano particles make up less than 5% or so of the insulation material by weight). And researchers needed to understand why if the particles were larger (in the more common micrometer scale), the voltage endurance improvement essentially disappeared. More recently research has concentrated on ensuring the nano particles are evenly dispersed – that is they do not agglomerate together to make a micro‐sized particle, where the voltage endurance properties are lost. This latter aspect has been a great challenge, and involves consider‐able proprietary technology. It is probably why nanodielectrics have not been widely applied on a commercial scale to date.
The first application of nano technology in the rotating machine field was as magnet wire insulation for inverter duty (variable speed) motor stator windings. Invertors of the voltage source, pulse width modulated type generate short risetime voltage surges that applied a high voltage to the turn–to‐turn insulation.
Even 440 V motors were found to have PD in small voids between the magnet wires due to these voltage surges. Since the magnet wire film insulation is usually purely organic, it had very little PD resistance, and winding failure occurred quickly. Nanodioelectrics, a natural for this application, imparting much longer life due to its much higher PD resistance.
However, application of nanodielectrics to the main groundwall insulation of high voltage stators is still not commercial. Unlike the very thin film on magnet wire, the groundwall of a high voltage coil is several mm thick, and getting good dispersal of the nanodielectrics resin is probably the main challenge. Recently Siemens published work they have been doing to introduce a nano particle filled ep‐oxy as an impregnating material in VPI bars and coils [3]. What is surprising is that they indicate the filler may be about 25% by weight, which should make for a viscous VPI impregnation material. At this stage, the nanodielectric is not used to replace the mica paper tape, but instead is just used to bond the tapes together with a much more PD‐resistant material than pure epoxy. Siemens is now making full scale coils to evaluate the new material (and the processing required). If the new groundwall insulation demonstrates the huge in‐crease in voltage endurance that small experimental coils yielded ‐ and assuming the cost is reasonable – we will likely see in the near future the big‐gest change in motor and generator insulation technology in over 50 years.
References
1. C. Stone, G.H. Miller, “Progress in Rotating Ma‐chine Insulation Systems and Processing”, IEEE Electrical Insulation Magazine, July 2013, pp 45‐51.
2. IEEE Electrical Insulation Magazine, November 2013 Issue.
3. Gropper et al, “Nanotechnology in high voltage insulation systems for large electrical machinery – first results”, CIGRE Paper A1‐103, August 2012.