The calculation of the output current of distribution generation (DG) units interfaced with an inverter during the fault is a major issue for isolated and grid-connected distributed networks. The droop control inverter interfaced DG has controlled output current within 2 pu. during the fault. Furthermore, the current output of DG during the fault depends on solar irradiation and wind speed, increasing the uncertainty due to the intermittent nature of renewable energy sources. The installation of DG modifies the fault current direction and strength, which makes relay coordination more difficult. Overcurrent relays are used to defend isolated and grid-connected microgrids. This paper uses different techniques to study the fault current's probability distribution function (PDF) for isolated and grid-connected MGs. We use the droop control and virtual impedance techniques to calculate the probability of the short circuit current that the inverter-interfaced DG contributes. Wind and PV system output power generation samples are tested on MGs using the Monte Carlo Simulation (MCS) approach. A coordination time probability for relays on each line has been calculated to find the mean and standard deviation values of a setting time for overcurrent relays on a faulted bus. The proposed probabilistic model has been tested on the isolated and grid-connected IEEE 33-bus with 5 DGs and MGs using MATLAB code. We found that the droop control method gives a much longer overcurrent relay operating time than the virtual impedance method. This is true for DG buses for both modes of isolated and grid-connected MGs, as well as buses that connect branches. Additionally, for two modes—isolated and grid-connected MGs—the standard deviation of the relay operating time calculated by droop control is higher than its value calculated by the virtual impedance on the same bus.