Ultra-high voltage (>30KV) power devices through superior materials for HVDC transmission

点击翻译按钮获取中文摘要

In October 2014 the UK energy surplus during winter months dropped to below 5% overcapacity. In the future, this emergency overcapacity may be further diminished and actually become devastatingly insufficient, necessitating the national grid to divert large power demands to areas at opposite ends of the country or face serious and harmful disruption to energy supply. A viable solution to this supply problem is to build new national high voltage DC (HVDC) energy network connections in addition to more international connections to the super grid. To implement HVDC effectively, companies are considering two options: 1) to implement mature Si 300 MW HVDC technologies (circa 2009) requiring large overhead, land requirements, maintenance costs and cooling systems by scaling with the number of converters per line. Or 2) to invest in technologies which upscale the blocking voltage and the current capacity of individual power devices in a converter where fewer line converters and greater efficiency can be achieved for 2 GW MMC HVDC. Even by reducing the series chain effect in power transmission across the UK and conversion a 3% saving can equate to three 500 MW coal power stations from the current UK power usage of approximately 37 GW.This fellowship seeks to develop revolutionary Silicon Carbide (SiC) material for ultra-high voltage (UHV) >30 kV power devices with large current ratings, up to 150A, with the intention of pushing the current rating as far as possible. The current rating of UHV vertical devices depends on availability of large surface areas (> 1 cm2) and is presently limited due to defects from excess material deposits forming on the wafer during the material growth. This is a problem which I believe will be a critical roadblock to such technology and receives little attention as the proposed power ratings are currently off the 5-10 year power electronics roadmap. Problematically, many in the field trust it will be solved at that point, however, no major research drive is currently underway to solve this essential problem.Chemical vapour deposition (CVD) is the industry gold-standard technique for creating the semiconductor materials used in these UHV devices due to its excellent uniformity, scalability and reproducibility and so must be developed for quick uptake of any power device technology. For UHV devices the material choice and quality is key, where a defect free, thick (~100 um) layer with a large surface area is needed. Chlorinated chemistry is a recent development in SiC CVD and helps push growth rates up to 100 um/hr for thick layers and can be used to better achieve low background doping densities which are both required for high power technology. Here, thick high quality material will be achieved by state of the art epitaxial growth in the UK's only industrial SiC CVD at Warwick. In tandem to improving the material, its superiority will be shown by fabrication of vertical UHV devices: Schottky Diodes, PiN Diodes and MOSFETS, whilst developing IGBT processing, to show their potential for future modular multi-level converter (MMC) HVDC networks.Bipolar devices such as Si thyristors and Si PiN diodes will be used in rail traction and grid level HVDC applications due to the high quality of Si material which allows large current ratings. In 2014, Yole predicted that these sectors would boost the >3.3 kV SiC market and in 2015 Mitsubishi showed an all SiC 3.3 kV traction inverter system. I further predict that SiC devices will only completely replace Si IGBTs in the > 10 kV range when the current-limiting surface defects are minimised and device reliability due to minor material defects is better understood. Only then will large current ratings be achieved, which will allow the technology to surpass current HVDC technology. This fellowship directly studies these limiting mechanisms and will develop the material and associated technology to underpin this step change in power technology