We consider the problem of energy-efficient base-station (BS) planning for green cellular network design. There exist a number of criteria for greenness in the literature, but we focus only on the energy-normalized throughput. We first model energy consumption for a heterogeneous network, consisting of macro and micro base-stations, and a network topology. Then, we propose an iterative algorithm for the green BS planning problem in the one-dimensional case: 1) We find the positions of base-stations that maximize the energy-normalized throughput for a fixed number of base-stations, and 2) find the optimal number of base-stations which maximizes the energy-normalized throughput. In this work, we evaluate the energy-normalized throughput accounting for adaptive modulation rates with power control, which is a common feature of modern communications systems. The convergence to a local maximum for the proposed algorithm is shown using computer simulations, and the corresponding energy-normalized throughputs are evaluated for a number of system model parameters.