The exponential growth of unpredictable renewable power production sources in the power grid results in hard to regulate reactive power. The ultimate goal of ORPD is to compute the optimal voltage level of all the generators except reference bus, off-nominal turns ratio of transformer and MVAR injection of shunt var compensators (SVC) values. More realistically, ORPD problem is a multi-objective problem. Therefore, in this paper, s minimization of active power loss, voltage deviation and operating cost of renewable and thermal generators are considered the objective functions (formulation of three cases of two and three objective functions). Usually, renewable power generators such as wind and solar and load demand are uncertain. Therefore, probabilistic mathematical modeling such as normal, Weibull and lognormal probability distribution functions (PDFs) are implemented to model the generation and demand. to generate 1000 scenarios with the help of Monte-Carlo simulation (MCS) techniques. Afterward, to reduce the computational burden, scenario reduction technique is applied to pick 24 representative scenarios. These 24 scenarios are solved by using constrained coevolutionary multi-objective optimization (CCMO) algorithm. IEEE 30 bus test system is considered to achieve effectiveness and superiority of CCMO. Three stochastic study cases have been analyzed in the simulation results. Simulation results indicate that the proposed algorithm is used to detect the global optimal solution of ORPD problem.