Nuclear thermal-hydraulic system analysis codes with the Best Estimate plus Uncertainty approach are unavoidably uncertain and naturally, an uncertainty analysis has been an important issue. In general, the system is discretized depending on the user while satisfying some criteria. Previously, uncertainties related to spatial discretization in nodalization have been less considered due to their difficulty in quantification. However, there are some results that they are comparable to other uncertainties. Previous studies have been conducted by evaluating and optimizing the uncertainty related to user effect by performing iterative calculations. Since these methods consume a large amount of resources for computation, this study suggests a method of consuming less resources. For this purpose, the adjoint method is applied to the code for node optimization in this study. The proposed methodology is applied to two cases and confirmed its applicability for the node optimization and its utility in the node uncertainty quantification.