In this paper, we propose to maximize the energy efficiency (EE) of a two-tier heterogeneous network (HetNet) by jointly optimizing the activation probabilities of base stations (BSs) and a partial spectrum reuse (PSR) scheme where small cells are allowed to reuse a portion of the spectrum of the macrocells. First, we obtain the coverage probabilities of macro cell users (MUs) and small cell users (SUs) as functions of the probability of a small base station (SBS) being active, the probability of a macro base station (MBS) being active and the PSR factor. Then, we formulate the EE maximization problem based on the coverage probabilities. We obtain the optimal values of the activation probabilities of an SBS and an MBS as well as the PSR factor by solving the EE maximization problem using Genetic Algorithm (GA). Comparing with an existing work, the optimization relies on two factors (i.e., activation probability ratio and partial spectrum reuse factor) while our work depends on three combination factors (i.e., activation probability of MBS, activation probability of SBS and PSR factor) that leads to a higher EE. Numerical results show that a higher activation probability of an SBS and a higher PSR factor lead to a higher EE of the HetNet, while a higher activation probability of an MBS leads to a lower EE of the HetNet.