This study presents the rotor blade airfoil analysis of residential-scale wind turbines. On this track, four new airfoils (GOE 447, GOE 446, NACA 6412, and NACA 64(3)-618) characterized by their high lift-to-drag ratios (161.3, 148.7, 142.7, and 136.3, respectively). These new airfoils are used to generate an entire 7 m long blades for three-bladed rotor horizontal axis wind turbine models tested numerically at low, medium, and rated wind speeds of 7.5, 10, and 12.5 m/s, respectively, with a design tip speed ratio of 7. The criterion to judge each model’s performance is power output. Thus, the blades of the model that produce the highest power are selected to undergo a tip modification (winglet) and leading-edge modification (tubercles), seeking power improvement. It is found that the GOE 447 airfoil outperformed the other three airfoils at all tested wind speeds. Thus, it is opted for adding winglets and tubercles. At 12.5 m/s, winglet design produced 5% more power, while tubercles produced 5.5% more power than the GOE 447 baseline design. Furthermore, the computational domain is divided into two regions: rotating (the disc that encloses the rotor) and stationary (the rest of the flow domain). Meanwhile, the numerical model is validated against the experimental velocity measurements. Since Reynolds-averaged Navier–Stokes with k–ω shear stress transport turbulence model can capture the laminar-to-turbulent boundary layer transition, it is used in the 18 simulations of the current work. However, large eddy simulation (LES) can deal successfully with the various scale eddies resulting from the rotor blades and its interactions with the surrounding flow. Thus, the LES was used in the six simulations done at the rated wind speed. LES power output calculation is 7.9% to 11.9% higher than the RANS power output calculation.