With all the emphasis on inverter efficiency, why is the physical limit to solar panel conversion efficiency (solar insolation to electrical power) is around 30% with current technology? At the moment, the maximum reached in a lab is about 26.5% and production panel efficiency typically ranges from 17 to 21%, with potential improvements centered around careful mechanical design to minimize ‘cell shading’ and to collect every possible photon available. Why not make the inverter cheaper, although somewhat less efficient?
Power loss creates heat that must be dealt with. Consider a 1 % improvement in efficiency of a 500 kVA inverter system. This is tantamount to turning off five 1 kW space heaters inside the already heat-soaked cabinet and power electronics is all about heat. The problem is it is low-quality heat, not easily recaptured for cogeneration of energy. Reducing heat reduces heat sink cost while simultaneously improving reliability. Increasing efficiency however adds cost regarding the number and type of power semiconductors, and there is definitely a point of diminishing returns.
The reality is that some of the efficiency gains from SiC devices are “spent” on increasing the switching frequency, which reduces the size and cost of magnetics and capacitors, thus offsetting the higher cost of SiC with system-level cost savings. The good news is that the switching frequency can increased substantially over that of IGBTs, certainly above the audible range, with only a tiny increase in total power loss.