In this paper, an analytical framework is provided to analyze the energy coverage performance of unmanned aerial vehicle (UAV) energy harvesting networks with clustered user equipments (UEs). Locations of UAVs are modeled as a Poison Point Process (PPP), while locations of UEs are modeled as a Poisson Cluster Process (PCP). Two different models are considered for the line-of-sight (LOS) probability function to compare their effect on the network performance. Moreover, ultra-wideband (UWB) antennas with doughnut-shaped radiation patterns are employed in both UAVs and UEs, and the impact of practical 3D antenna radiation patterns on the network performance is also investigated. Initially, the complementary cumulative distribution function (CCDF) and probability density function (PDF) of path losses for each tier are derived. Subsequently, association probabilities with each tier are obtained. Energy coverage probability is derived for the entire network using tools from stochastic geometry. Via numerical results, we show that UAV height and antenna orientation play crucial roles on the energy coverage performance.