Turbulent Rayleigh-Benard convection is ubiquitous in nature and many engineering applications. To better understand the heat transfer physics in turbulent Rayleigh-Benard convection, researchers typically perform flow visualization studies. However, implementing flow visualization techniques in an experimental setting necessitates the removal of insulation that runs over the periphery of the sidewalls enclosing the convecting fluid within the convection test cell. Although temperature of the bulk of convecting fluid within the test cell and the ambient temperature (temperature outside the test cell) may be nearly identical, there can still be a substantial effect due to this uninsulated, relatively high-thermal conductivity, transparent sidewalls on the measured Nusselt number. The present study aims to combat this discrepancy in Nusselt number through a semi-analytical model based on the concept of extended surfaces. 2-D numerical simulations of this scenario are performed using ANSYS Fluent for a wide range of Rayleigh numbers (10⁶ - 10¹¹) at a fixed aspect ratio of unity, while keeping a constant temperature difference between the isothermal plates (8K), to determine some of the unknown variables in the proposed semi-analytical model. Relations needed to estimate these unknown variables are provided to facilitate ease of use for future researchers when applying this model to determine their sidewall corrections. Sidewall-corrected Nusselt numbers obtained using the proposed semi-analytical model are then compared with those estimated using an existing analytical model from the literature (which is only applicable for R-B convection systems having sidewalls with perfect external insulation). This comparison reveals a significant difference in the estimated sidewall corrections (~ 64%), suggesting that the model proposed for correcting thermal conductance loss through externally-insulated sidewalls is inappropriate and inadequate when correcting for thermal conductance heat loss through uninsulated sidewalls.