Abstract The Colombian electrical power system is being transformed by the large-scale integration of energy storage systems and renewable energy resources to the power system. These technologies can be integrated using alternating current or direct current technologies via power electronic converters. Here we analyze the direct current paradigm by proposing a master-slave optimizer for solving the problem of optimal power flow considering nonlinear loads. The master stage covers the dispatch of all distributed generators using the hurricane optimization algorithm. In the slave stage, a power flow method based on successive approximations is used to determine the voltage variables and evaluate the objective function of the problem, which is the minimization of power loss. A 69-nodes direct current network is used as a test case to compare the numerical performance of the hurricane optimization algorithm with a nonlinear optimization package and a metaheuristic approach called black hole optimizer. All simulations are performed using MATLAB software version 2017 a licensed by Universidad Tecnológica de Bolívar, Colombia.