Wireless power transfer may enable remotely powered operation of low-energy devices in challenging environments such as tactical networks. Millimeter wave (mmWave) is a possible candidate for wireless power transfer, thanks to the use of directional antenna arrays. This letter presents a feasibility study of using mmWave for wireless power transfer in a large-scale network consisting of power beacons and energy harvesters. Using stochastic geometry, system performance is characterized while treating the network nodes as potential blockages to mmWave signals. A link-level metric (energy coverage probability) and a network-level metric (overall success probability) are considered. The former captures whether a typical energy transfer link provides the requisite energy, while the latter also takes the network load into account. Numerical simulations suggest that network densification helps improve the performance, despite an increase in the blockage density. For a given density, deploying an optimal fraction of nodes as power beacons maximizes the number of successful energy harvesters. Finally, mmWave outperforms lower frequency solutions despite blockages.