With the increasing participation of renewable energy sources in the production of electricity, the answer to the question of the necessary operating reserve capacity in power systems is becoming more and more complicated. The very fact of the variability of driving energy (sun, wind, water) in photovoltaic power plants, wind power plants and small hydropower plants makes the required operating reserve capacity variable. As the electric power system itself is stochastic, the inclusion of renewable energy sources in the operation of electric power systems emphasizes its stochastic characteristic even more. For this reason, the need for operating reserve in electric power systems from the aspect of renewable energy sources and with them in connection with micro-grids, is best solved using the tools of probability, that is, statistics and estimation. The issue of operating reserve in the power system was investigated by Bayesian statistics, which connects the simultaneous occurrence of a random variable (available capacity of renewable energy sources in micro-grids) with a statistical variable (load in micro-grids). The result of this research is the probability of a deficit or surplus of available energy in micro-grids, and the consequence of that probability is the probability of the need and degree of engagement of the operating reserve in the power system using system’s thermal electric power plants and/or hydro electric power plants.