AutoPas is a particle simulation library that aims to reduce the time to solution of molecular dynamics simulations by automatically choosing the most appropriate strategy for a given scenario. With the rising cost of energy, however, it becomes more and more important to consider not only computation time but also the energy consumed in the process. Modern processors provide interfaces to monitor their energy consumption. In this project, we analyze the relationship between time to solution and energy consumption, first in a simple-to-understand problem, and then in a molecular dynamics simulation. We show that the fastest algorithm is not necessarily the most energy efficient. We also extend AutoPas by the ability to optimize for energy consumption instead of time to completion.