Reactive power load analysis is essential to the efficient design and operation of power systems. This variable is directly related to the quality, stability, and efficiency of energy supply, especially in the context of the increasing integration of distributed energy resources. From this perspective, the generation of synthetic load profiles that replicate the statistical properties of real data is becoming a useful tool for simulation, impact assessment, and optimization of electrical networks. This research investigates the convolutional NICE generative model applicability in generating synthetic reactive power load profiles. The proposed methodology is based on the analysis of a set of real data from a smart campus. By applying the generative model, the aim is to replicate the intrinsic statistical characteristics of the real data, such as mean and standard deviation to generate synthetic load profiles. The results show that the model can generate synthetic curves with high statistical congruence with the original data, as validated by metrics such as Kullback-Leibler divergence, Jensen-Shannon divergence, and Wasserstein distance. Such metrics indicate satisfactory agreement between the distributions of real and synthetic data. This study contributes to the literature by exploring the generation of synthetic reactive power profiles, thereby extending the applicability of generative models to a new domain within energy load analysis.
