The aim of this paper is to present a simulation of distribution systems adding solar PV generation using Matlab and OpenDSS. A chronological Monte Carlo simulation was developed to determine the probabilistic three-phase power flow with statistical power demand and solar resource data. A probabilistic load model was also proposed to generate power samples using patterns from real measurements. The irradiance, temperature and load values for each time step are managed on Matlab and sent to OpenDSS for the calculation of three phase power flow. The model was tested with the Comprehensive Test Feeder (CTF) for different amount of realizations presenting convergent values after 2000 realizations. The circuit of the CTF was tested obtaining a voltage profile and three phase power flow with a maximum error of 2.78% compared with the original results. The CTF was modified adding 4 PV generation systems to low-voltage nodes with a 10.16% of rated power penetration with respect to the original load. The simulation allows to analyze the power flows of the feeder with PV generation and further variables by adding modules to the framework.