An oscillating water column (OWC) wave energy device can use a bidirectional flow impulse turbine for energy extraction. The airflow velocity profiles during inhalation and exhalation of the OWC are asymmetric, and the flowrate at exhalation is higher. Hence, this article presents an optimized turbine to work for asymmetrical airflow conditions. The rotor blade setting angle (γ) was made asymmetric with a non-zero setting angle, and the guide vane shape was optimized. The unsteady Reynolds-averaged Navier-Stokes equations (URANS) were solved to get the objective function values for the Latin hypercube sampling (LHS) generated design space. A pseudo-sinusoidal velocity profile replicating the flow inside the OWC device was used for the transient simulations. A weighted average surrogate model generated the initial population for a hybrid genetic algorithm to produce optimum designs. The effect of turbine induced damping and starting characteristics were analyzed under irregular flow conditions obtained from the Pierson–Moskowitz wave spectrum. The optimized turbine resulted in better starting and running characteristics. Also, the mean efficiency of the optimized turbine was increased relatively by 9.5%.