This study examines the performance and wake dynamics of vertical-axis hydrokinetic turbines (VAHTs) at low Reynolds numbers through an integrated computational approach, combining overset mesh and volume of fluid methods. Two- and three-dimensional simulations were conducted to evaluate mesh convergence, wake recovery, and power coefficient (Cp) under varying submersion depths. The results reveal that shallow submersion prolongs wake deficits, while turbines at greater depths exhibit faster wake recovery, aligning closely with experimental data. Additionally, the study demonstrates that even when free-surface deformation is minimal, the free surface significantly affects wake recovery, highlighting the need to account for free-surface effects. Furthermore, the study identifies turbulence model limitations, with overestimated wake deficits in the near wake, yielding a root mean square error (RMSE) of 0.1553 for the three-dimensional case compared to 0.4029 for the two-dimensional model, emphasizing the need of three-dimensional simulations for accurate wake recovery predictions. These findings highlight the importance of accounting for free-surface effects in VAHT placement and design, particularly in shallow water environments, to enhance performance and guide turbine deployment in tandem configurations. This work contributes to the broader effort of expanding decentralized renewable energy solutions for rural and off-grid communities in developing countries.