News Article
Electric vehicle go-slow hits SiC power devices
Wide-bandgap transistor and diode growth potential is obscured by automotive qualification, explains Yole Développement’s Philippe Roussel, though PV inverter growth is a bright spot.
Uncertainty hangs over the market for power devices made with wide-bandgap semiconductor SiC, due to a lack of clarity over whether and when electric vehicles will adopt them. “We have no firm estimate of when it will come,” said Philippe Roussel, business unit manager compound semiconductors, power electronics, LED & photovoltaics at market analysts Yole Développement in Lyon, France. “It’s still questionable.”
Automotive qualification can take up to five years, Roussel pointed out. So even if qualification for use in electric and hybrid electric vehicles (EV/HEV) is on-going it likely won’t be clear if SiC has been successful until 2015. And though qualification for EV/HEV charger inverters would be quicker, SiC faces a greater challenge there from silicon superjunction MOSFETs, IGBTs, and also wide-bandgap GaN devices.
As a result, when Yole releases its latest analysis on the SiC industry on May 8, it will describe two scenarios for SiC industry evolution. Its more optimistic scenario will see SiC devices used commercially in EV/HEV from 2015 onwards, taking 11% of the market from silicon IGBT devices by 2020. In the pessimistic scenario EV/HEV implementation doesn’t start until 2017/2018, making PV inverters the number one SiC application in 2020.
Today there remains much room for increasing SiC device usage in PV inverter applications, Roussel underlined. “Each inverter manufacturer’s product line-up has just one or two models with SiC in them, among dozens,” he said. “But it’s a very positive starting point.” That helped SiC power device industry revenues to $76 million in 2012, including R&D but excluding military use. PV inverter producers are the second industry to broadly adopt SiC devices, after manufacturers using SiC for power factor correction in high-end server power supplies.
Some PV inverter manufacturers use a SiC diode and silicon IGBT or MOSFET, and some offer full SiC inverters. “They’re just the first attempt, as inverter makers have limited their SiC development investment,” Roussel said. “They’re doing a simple replacement for silicon devices, taking the minimum extra work.” But that means that inverter producers are selling more expensive SiC products on their efficiency, without fully exploiting the material’s benefits. “The next step will definitely be a full redesign within the inverter that should fit with SiC’s high frequency and temperature capabilities, reducing the number of capacitors and inductors needed,” Roussel said.
Working with reverse costing specialist System Plus Consulting, Yole has modelled the benefits of increasing the standard PV inverter switching frequency from 12 to 32 kHz. That shortens the payback time on the SiC investment, and would make large 50 kW SiC inverters cheaper than their silicon equivalents by 2020. Such benefits will help increase annual revenues for SiC devices sold into PV inverters to $200 million by 2020, Yole predicts.
Even though PV inverters are currently better established, EV/HEV inverter producers could still be more advanced in making full use of SiC. “Efficiency drives adoption, but putting SiC in any system could also make it smaller and lighter,” Roussel stressed. “For EV/HEV, that is just fantastic.”
