The growing integration of rooftop photovoltaics (PVs) and energy storage units (ESUs) in customer households has resulted in changes in the customer load profiles. This is likely to influence the accuracy of state estimation (SE) carried out based on previously assumed load profiles. In this paper, a statistical model for modern low voltage (LV) customers was developed using Gaussian mixture model (GMM). The resulting model was subsequently applied to SE using weighted least squares (WLS) algorithm. LV network with high penetration of customer-owned PV and ESUs have been simulated. Different scenarios which include load profiles: with PVs integrated but without ESUs, ESUs alone, and with hybrid systems (combination of PVs and ESUs) have been considered. The results are presented and discussed.