Uncertainty management is becoming a very challenging tool in operation scheduling of networks interested by massive distributed generators' penetration. From this point of view, self-validated computing techniques are very useful tools, allowing to intrinsically track data uncertainty effects into power system operation procedures. In this field, the use of Affine Arithmetic has been demonstrated to be one of the most promising research direction, since it prevents the typical error explosion phenomena affecting the standard range arithmetic-based computing frameworks. Starting from these considerations, the aim of this paper is to investigate the Affine Arithmetic-based Power Flow problem, which, by mean of the uncertainty bounds characterization, can provide additional information to the system operators in their decision making processes. The methodology will be tested on the 30-bus IEEE test network and the results will be validated by application of traditional sampling-based techniques.