Abstract
The duck curve is a power demand curve that is named for its shape, and represents the demand on utility power generation. Recently, the peaks and valleys of the duck curve have been diverging. As more solar power is generated, mid-day supply becomes more abundant, resulting in decreasing simultaneous demand on utilities. These sudden changes in power availability can be damaging to the electrical infrastructure, e.g. by desensitizing safety equipment or tripping breakers. By modeling these loads on a computer and using predictive information on solar power availability, power loads can be scheduled to more closely match supply. Scheduling power loads to times when power is most abundant would shift those loads away from peak demand, both flattening the duck curve and reducing risk to the utility infrastructure.