Optimal control of multirotor drones is important due to the limited energy stored onboard such aircraft. This work presents a power model based on physical principles that capture the key system trends while being computationally simple. The model is validated by comparison to theoretical, qualitative and empirical results. Using the proposed model, energy-optimal allocation strategies are analyzed and compared to the widely used allocation matrix, which turns out to perform identically, for all practical purposes, in terms of power. By computing the power characteristics, various factors such as flight speed, orientation and drone weight are analyzed in terms of their influence on drone power consumption. For instance, by flying at a forward velocity of around 8 m s⁻1, one can increase the airtime of a typical drone by approximately 13 % compared to hover.